Skip to main content

Full text of "Woodland caribou restoration at Isle Royale National Park : A feasibility study"

See other formats


Woodland Caribou Restoration 
at Isle Royale National Park 

A Feasibility Study 

Jean Fitts Cochrane 
Technical Report NPS/NRISRO/NRTR/96-03 



1 







**m 







' 


^HV^I^^SpWL^ 


.i^. 





United States Department of the Interior • National Park Service 
Great Lakes Cluster • Isle Royale National Park 



The National Park Service disseminates the results of biological, physical, and social science research 
through the Natural Resources Technical Report Series. Natural resources inventories and monitoring 
activities, scientific literature reviews, bibliographies, and proceedings of technical workshops and confer- 
ences are also disseminated through this series. Documents in this series usually contain information of a 
preliminary nature and are prepared primarily for internal use within the National Park Service. This 
information is not intended for use in the open literature. 

This report comprises work completed in December 1992. 

Mention of trade names or commercial products does not constitute endorsement or recommendation for 
use by the National Park Service. 

The cover illustration is a photograph of the painting. Caribou at Middle Islands Passage, Isle Royale, by 
Howard Sivertson, Box 397, Grand Marais, MN 55604. 

Copies of this report are available from the author: 

Jean Cochrane 

Conservation Biology Program 

University of Minnesota 

1 00 Ecology Building 

1987 Upper Buford Drive 

St. Paul, MN 55108 



Woodland Caribou Restoration 
at Isle Royale National Park 

A Feasibility Study 



Jean Fitts Cochrane 

Isle Royale National Park 

800 East Lakeshore Drive 

Houghton, Ml 49931 



Present address: 

Conservation Biology Program 

University of Minnesota 

100 Ecology Building 

1987 Upper Buford Drive 

St. PauLMN 55108 



Technical Report NPS/NRISRO/NRTR/96-03 



September 1996 

United States Department of the Interior 

National Park Service 

Natural Resource Program Center 

Natural Resource Information Division 

Denver, Colorado 



Digitized by the Internet Archive 

in 2012 with funding from 

LYRASIS Members and Sloan Foundation 



http://archive.org/details/woodlandcaribourOOfitt 



Epigraph 



"Here [at the Island Mine spring in August of 1918 
or 1919] is where I first met the Isle Royale cari- 
bou. I had found their tracks at two or three points 
along the trail that morning and only a few minutes 
before had noticed that the juice was still flowing 
from a bitten milkweed. Then, right at 
the top of this rise, I rounded this little 
curve, and down there, where that 
rock lies by the road, some eight or 
ten rods away, stood three splendid 
young bulls, looking back to learn 
what followed them. Probably they 
had never seen a man before; at least 
they had no slightest fear. I stood for 
five minutes watching them and then 
walking forword [sic] until I had cut 
the distance in half before they turned 
and trotted over the little rise ahead. 
When I had climbed that, they were 
again looking back and waiting for 
me. To see what they would do, I 
yelled, but even then they ran away 
slowly, though I saw nothing more of 
them." 



whether there are seven or nine calves. It may be 
somebody has seen a better section of the Isle 
Royale caribou herd, but I have never heard them 
tell of it." 




"[And then watching from the beach 
at Lake Desor, after a sunrise 
swim]. ... A half mile away, headed 
from the largest island, where they have been 
spending the night, perhaps to keep the calves 
secure from attack by the wolves 1 ", a line of caribou 
is swimming toward the main shore. Noses up, 
antlers lying back almost upon their shoulders (the 
cows have antlers as well as the bulls) with the 
calves swimming bravely alongside their mothers, 
they make a swift passage; and we see them climb 
the bank, their dripping forms sharply outlined 
against the green background. We count a full 
baker's dozen of the adults and are a little in doubt 



Woodland caribou swimming, Slate Islands, Ontario, Canada. 
Caribou are excellent swimmers and regularly swim between islands 
to seek refuge from predators. Photograph: H.R. Timmermann, 
Ontario Ministry of Natural Resources. 



— W.RF. Ferguson, "Wildlife and Adventures 

Along the Great Desor Trail on Isle Royale," 

Detroit News, 28 May 1922 



1 A popular term for the coyote (Cants latrans) in those days was "brush wolf.' 



Contents 



List of Figures \ vii 

List of Tables \ vii 

Acknowledgments \ ix 

Summary \ xi 

Introduction \ 1 

Study Objectives \ 2 
Background \ 2 

Isle Roy ale National Park \ 2 

Caribou Distribution and Decline in the Lake Superior Region \ 5 

Caribou on Isle Royale \ 1 

Methods \ 9 

Results and Discussion \ 1 1 

Why Did Caribou Disappear From Isle Royale? \ 1 1 

Migration \ 1 1 

Disease \ 1 1 

Competition with Moose \ 1 2 

Predation \ 13 

Overhunting \ 14 

Fires \ 14 

Summary \ 1 5 
Habitat Suitability \ 15 

Forage Resources \ 15 

Predators \ 21 

Winter Ticks \ 22 

Escape Habitat \ 23 

Summary \ 26 
Population Vulnerability Analysis \ 27 
VORTEX Results \ 27 

Lessons From Elsewhere \ 28 

Summary \ 29 
Reestablishment Methods \ 29 

Alternative Reestablishment Strategies \ 30 

Release Timing \ 30 

Disease and Parasite Screening \ 30 

Sources of Caribou — the Slate Islands \ 31 

Summary \ 33 



Impacts and Management Implications \ 33 
Vegetation \ 33 

Endangered Species and Rare Plant Communities \ 33 
Other Animals \ 34 
Diseases and Parasites \ 34 
Wilderness \ 35 
Summary \ 36 

Conclusions \ 39 

Literature Cited \ 43 

Postscript \ 53 

Appendix A: Summary of Caribou and Selected Reindeer Translocation Attempts \ 55 
Literature Cited \ 57 

Appendix B: VORTEX: Simulation Model of Stochastic Population Change \ 59 
Stochastic Simulation of Population Extinction \ 59 
Literature Cited \ 61 

Appendix C: The History of Caribou in the Lake Superior Region \ 63 
Mainland Michigan (Excluding Isle Royale) \ 63 
Northeastern Minnesota \ 63 
Northwestern Ontario \ 65 
Isle Royale \ 67 

Appendix D: Caribou Ecology \ 73 
Habitat Use \ 73 
Home Range and Density \ 75 
Escape Habitat and Predator Densities \ 76 
Diet \ 77 
Population Dynamics \ 77 

Appendix E: Summary of VORTEX Modeling Runs \ 79 



Contents 



Figures 



Epigraph — Woodland caribou swimming \ iii 

1. Location of Isle Royale National Park \ 2 

2. Topography of Isle Royale \ 3 

3. Moose and wolf populations on Isle Royale \4 

4. Isle Royale National Park map \ 5 

5. Range limits of caribou in the Great Lakes region \ 6 

6. Relict and introduced caribou around Lake Superior \ 7 

7. Woodland caribou crossing a frozen lake \ 11 

8. Caribou hunters on north shore of Lake Superior \ 14 

9. Vegetation type map of Isle Royale National Park \ 1 6 

10. Savanna habitat near Windigo created by intensive moose browse \ 

11. Lightly browsed habitat on Smithwick Island \ 18 

12. Bog on Raspberry Island \ 19 

13. Jack pine stand with ground lichens, Saginaw Peninsula \ 19 

14. Spruce-fir forest on Mott Island \ 20 

15. A pair of wolves on shore ice, Isle Royale \ 21 

16. Major offshore islet groups, Isle Royale \ 24 

17. Rock Harbor channel in winter with extensive ice cover \ 25 

1 8. Capturing woodland caribou on the Slate Islands \ 32 

19. Preparing caribou for transport off the Slate Islands \ 32 

Postscript — Female woodland caribou \ 53 



Tables 



1. Isle Royale vegetation and nonvegetation cover types and areas covered \ 17 

2. Offshore islets, Isle Royale National Park \ 24 

3. Simulated population persistence times \ 28 

4. Population values in VORTEX simulations \ 80 

5. Mortality rates by sex and age class \ 81 

6. Woodland caribou population simulations run in VORTEX \ 82 

7. Sample VORTEX results, run number 4 \ 83 



Contents 



Acknowledgments 



Numerous people aided this study immeasurably by 
discussing and sharing literature on caribou biology, 
caribou translocation, and Isle Royale's resources 
with me. Chris Martin's draft paper on the history 
of caribou on Isle Royale stimulated my interest in 
the project. Tim Cochrane contributed a wealth of 
historical leads and insights, especially concerning 
Native American use of Isle Royale and the island's 
relations to the mainland. Terry Kreeger's prelimi- 
nary VORTEX modeling runs were invaluable. 



Financial support for this feasibility study was 
provided by the North Central Caribou Corpora- 
tion, a nonprofit organization promoting woodland 
caribou restoration in the Minnesota-Ontario 
borderlands, and the Michigan Department of 
Natural Resources, Natural Heritage Program. Isle 
Royale National Park provided logistical support 
for field and office work. 



Although brief, a June 1990 visit to Isle Royale by 
caribou expert A.T (Tom) Bergerud of British 
Columbia and H.R. (Tim) Timmermann of Ontario 
was a great asset. Peter Gogan, Rolf Peterson, and 
Joanne Thurber assisted with study planning and 
discussions throughout the project. Finally, Bob 
Krumenaker, former park resource management 
specialist, was very supportive from the beginning 
of the study. Others who took time to share their 
knowledge were Dick Anderson, Tom Cooley, 
Bruce Connery, Gordon Eason, Jay Gore, Peter 
Jordan, Murray Lankester, Mark McCollough, Jim 
Nelson, Rick Page, Bill Samuels, Greg Servheen, 
Steve Schmidt, and additional members of the North 
Central Caribou Corporation. Thanks also to Mary 
Fitts, Nan Cochrane, Greta Pelto, Julie Cochrane, 
and Sue Salaniuk, and to Corrine, Andrew, and 
Madaline. 



Summary 



Interest in restoring woodland caribou (Rangifer 
tarandus caribou) to Isle Royale National Park was 
sparked by recent caribou relocations on other Lake 
Superior islands in Ontario and by the marked 
decline of wolf (Cards lupus) numbers on Isle 
Royale 1 . Further stimulus came from caribou 
biologists (Bergerud and Mercer 1989) suggesting 
that the absence of white-tailed deer (Odocoi tens 
virginianus) and predators other than wolves should 
favor caribou restoration on Isle Royale compared 
to most mainland sites. 

National Park Service (NPS) policies provide 
direction for the restoration of native species in 
National Parks. Criteria that must be evaluated 
include the species' historical presence and the role 
of humans in its extirpation, and project feasibility 
including the prognosis for establishing a self- 
sustaining, viable population without harming other 
resources or people. 

Woodland caribou roamed over Isle Royale for 
more than 3,500 years until they disappeared 
around 1928. Overhunting throughout the Lake 
Superior region in the 19th century, including on 
Isle Royale, initiated the species' decline. The 
caribou's regional extirpation has also been linked 
to the northward range expansion of white-tailed 
deer and the increase in moose (Alces alces) and 
wolf numbers that followed forest clearing by Euro- 
American settlers. While wolves and deer did not 
immigrate to Isle Royale before the caribou disap- 
peared, moose and coyotes (Canis latrans) arrived 
on the island shortly after the turn of the century. 
Coyote and native lynx (Lynx canadensis) predation 
likely eliminated the remnant caribou herd on Isle 
Royale, which by this time was isolated from the 



caribou population in northern Ontario. Thus, 
humans contributed both directly and indirectly to 
the caribou's decline on the island. 

The feasibility of restoring caribou to Isle Royale 
depends upon the suitability of habitats currently 
available in the park, the estimated vulnerability of 
a translocated population to extirpation, the logis- 
tics and costs of translocation and monitoring, and 
the likely impacts and management requirements of 
a restored caribou herd. Each of these topics is 
addressed in this study. 

In brief, forage resources on Isle Royale appear to 
be suitable for caribou. The key predictor of habitat 
quality, however, is the adequacy of escape habitat 2 
to buffer caribou from wolf predation. The numer- 
ous offshore islets surrounding the main island of 
Isle Royale would provide escape habitat, except in 
winters when ice forms between the islets and the 
main island. In some winters the weather may be 
severe enough to eliminate all water-related escape 
havens. We cannot predict whether wolves would 
learn to seek out caribou or whether introduced 
caribou would learn to use available escape habitat. 

Population modeling indicates that a small caribou 
herd would have a high probability (90%) of 
extirpation within 40-60 years of release. This 
prediction assumes that wolves would be present at 
high densities (0.01/km 2 ), restricting the caribou to 
secure habitats and limiting the population to 54 or, 
more likely, fewer caribou. Hence, a self-sustaining, 
viable population would not be achieved in isolation 
from other populations. 



1 In this text, "Isle Royale," "the island," and "the park" refer to the main island and surrounding archipelago now comprising 
Isle Royale National Park. 



2 "Escape habitat" means physical features in the landscape that allow caribou to avoid detection by predators, or elude 
predators if they are detected and pursued. Examples include islands, large bogs, and rugged terrain. 



Woodland caribou from the Slate Islands, Ontario, 
are available for translocation, duplicating proce- 
dures used successfully by the Ontario Ministry of 
Natural Resources. Disease and parasite screening 
and treatment for import into the United States 
could complicate the translocation process. The 
released animals would have to be monitored 
carefully for a number of years to assess the success 
of the restoration. Translocation and monitoring 
costs for a full-scale restoration project could reach 
$ 1 00,000 per year for three or more years. 

Potential impacts of a small caribou herd include 
threats to rare plants and introduction of diseases or 
parasites to Isle Royale's isolated moose herd. 
Although woodland caribou are a historic element 
of the park, the island ecosystem has continued to 
evolve in the decades since caribou disappeared. 
These concerns would have to be investigated as 
part of the environmental assessment of any caribou 
restoration plan. 

Isle Royale is a National Park and an island wilder- 
ness supporting unique floral and faunal commun- 
ities. In this setting, managers face special chal- 
lenges if they attempt to sustain or restore viable 
populations of large, rare animals, such as caribou 
and wolves, while maintaining other resource 
values. 



S u miliary 



Introduction 



Restoration of extirpated species has been promoted 
since the first investigations of wildlife populations 
in the National Parks (Wright et al. 1933). In 1932, 
Wright et al. (1933) called for "biological engineer- 
ing" to restore parks to their "primitive dynamic 
balance." Their proposed policy for the National 
Parks stated "that any native species which has 
been exterminated from the park area shall be 
brought back if this can be done. . . . Restoration of 
an animal which has been exterminated is desirable 
not only because it will bring back that species 
itself, but because it will fill once more the niche 
that was deserted. . . ." (ibid, p. 23). 

In 1963, the Leopold Committee Report (Leopold et 
al. 1963) called for restoring naturally functioning 
ecosystems, with the goal that national parks 
"represent a vignette of primitive America" before 
Euro- American settlement. In a subsequent report 
titled, "National Parks: From Vignettes to a Global 
View" (often called the Gordon Report after the 
principal author), the Commission on Research and 
Resource Management Policy in the National Park 
System ( 1 989) called for the National Park Service 
to maintain and restore native biota, but to resist 
establishing alien species. The Gordon Report urged 
NPS managers to become "ecological engineers" 
when necessary to keep ecosystems functioning 
(ibid). 

Under current management policies (U.S. Depart- 
ment of the Interior, National Park Servce 1988, 
1 99 1 ) the NPS will strive to restore native species 
to parks wherever all of the following criteria can 
be met: 

1 . Adequate proof exists that the species 

occurred in the area and that it disappeared, 
or was substantially diminished, as a direct 
or indirect result of human-induced change to 
the species population or to the ecosystem. 



2. Adequate habitat (land, water and other 
essential elements) to support the species 
either exists or can be reasonably restored in 
the park and if necessary on adjacent public 
lands and waters, and once a natural popu- 
lation level is achieved, it can be self- 
perpetuating (a "viable population"). 

3. The species does not pose a serious threat to 
the safety of park visitors, park resources, or 
persons or property outside park boundaries. 

4. There are no significant problems with 
predators at the release site, or the problems 
can be resolved. 

5. The subspecies used in restoration most 
nearly approximates the extirpated subspe- 



6. 



cies or race. 



A review indicates that the prospects for 
natural reestablishment are minimal, but that 
restoration has a good chance of success. 



7. A restoration action plan has been developed. 

The purpose of this study is to assess the feasibility 
of restoring woodland caribou within Isle Royale 
National Park and address the NPS restoration 
criteria. A previous study (Martin 1988) reviewed 
the history of caribou on Isle Royale and provided a 
preliminary assessment of restoration potential. 
Interest in reintroducing caribou to Isle Royale has 
heightened recently due to transplantation successes 
on other Lake Superior islands and the decline of 
wolf numbers on Isle Royale (Bergerud and Mercer 
1989, Cutler 1989). Bergerud and Mercer (1989) 
contended that, with no white-tailed deer, attempts 
at Isle Royale should be favorable compared to 
mainland sites infected with the meningeal brain- 
worm (Parastrongylits tenuis), a parasite fatal to 
woodland caribou. 



Study Objectives 

This feasibility study addresses five objectives: 

1 . Determine historical caribou population 
levels and the causes of their decline on Isle 
Royale. 



2. Assess the suitability of habitat on Isle 
Royale for sustaining a caribou population. 

3. Analyze the vulnerability of an introduced 
population to extirpation. 



Background 

Isle Royale National Park 

Isle Royale National Park is a 544-km 2 island 
archipelago lying 24 km south of the Canadian 
mainland in Lake Superior (Figure 1 ). The main 
island of Isle Royale stretches 72 km long and 14 
km wide and is surrounded by more than 200 
smaller islets. A series of parallel ridges and valleys 
running the length of the main island characterizes 
Isle Royale's topography, with steep, north-facing 
cliffs and gradual south slopes (Figure 2). Inland 
lakes and linear bogs are numerous. 



4. Assess the feasibility of translocating caribou Isle Royale's climate is cool and moist, with a 



to Isle Royale. 

Identify the potential impacts and manage- 
ment implications of reintroducing caribou 
on Isle Royale. 



notable moderating influence from Lake Superior 
(Hansen et al. 1973). Mean growing season tem- 
perature is 1 1 .6°C (Strommen 1969) and mean 
rainfall exceeds 5 cm every summer month (Slavik 
and Janke 1987). Snow falls from October to April 





Thunder 
Ba V#/ 


^ISLE ROYALE 


ONTARIO 


Duluth 




/^Houghton 

MICHIGAN 




MINNESOTA \ 


WISCONSIN 






• 








Twin Cities \^ 




Green [ / 
Bay / ) 





Figure I. Location of Isle Royale National Park in Lake Superior. 



Introduction 



or May, with midwinter snow depth 
usually 60-77 cm (Peterson and Allen 
1974; R.J. Krumenaker, pers. comm.). 
Ice frequently forms between the 
closest surrounding islets and the 
main island, and ice occasionally 
forms around all islets and to the 
Canadian mainland (R.O. Peterson, 
pers. comm.). 

Isle Royale falls at the southernmost 
limit of the northern coniferous forest 
region (Slavik and Janke 1987). The 
balsam fir, white spruce, paper birch, 
and quaking aspen (Abies balsamea, 
Picea glauca, Betula papyrifera, and 
Populus tremuloides) forest types 
cover nearly two-thirds of Isle Royale 
(Linn 1957, Hansen et al. 1973). At 
higher elevations away from the lakeshore, sugar 
maple and yellow birch (Acer saccharum and 
Betula alleghaniensis) forests dominate (Linn 
1957), but cover only about 5% of the island 
(Hansen et al. 1973). This forest community 
represents a northern limit of the temperate decidu- 
ous or Great Lakes-St. Lawrence forest region. 
Other less prevalent plant communities on Isle 
Royale include black spruce (Picea mariana) bogs, 
northern white cedar (Thuja occidentalis) swamps, 
rock outcrops, and jack pine (Pinus banksiana) and 
black spruce stands. In 1936, a fire fueled by 
logging slash and drought burned nearly 20% of Isle 
Royale, resulting in an almost pure stand of paper 
birch (Slavik and Janke 1987). Natural fire fre- 
quency is probably only 0.5-1 .0 fires per year 
(Hansen et al. 1973; R.J. Krumenaker, pers. 
comm.), and recent fires have demonstrated the 
typically slow-burning condition of forest fuels on 
Isle Royale. 

Characteristic of an isolated island, the fauna on 
Isle Royale is less diverse than an area of similar 
size on the adjoining mainland. At present, muskrat 
(Ondatra zibethicus), beaver (Castor canadensis), 
snowshoe hare (Lepus americanus), short-tailed 
weasel (Mustela erminea), mink (Mustela vison), 
river otter (Lutra canadensis), red fox (Vulpes 




Figure 2. Topography of Isle Royale. This aerial photograph of the 
northeast end of Isle Royale illustrates the island's ridge and valley 
topography, with a fingerlike shoreline of long bays ringed by chains 
of islets. Photograph: Isle Royale National Park. 



vulpes), timber wolf, and moose are the only larger 
mammals living on Isle Royale (Jordan 1981). 
Species composition has shifted markedly in this 
century, including the arrival of moose, red foxes, 
coyotes, and wolves, and the disappearance of 
marten (Martes americana), lynx (Lynx cana- 
densis), coyotes, and caribou (ibid). White-tailed 
deer were introduced in 1910 but never prospered; 
none were seen after 1936. 

In addition to reduced diversity, islands often 
support atypically high animal densities, and Isle 
Royale's moose and wolf populations fit this 
pattern. Moose arrived in 1905 and increased 
rapidly to 1,000-3,000 in the early 1930s, then the 
population crashed to about 200 animals by 1935 
(Krefting 1974). Forest regeneration after the 1936 
fire stimulated another less dramatic population 
surge followed by a die-off in 1 948-50. Since 
wolves immigrated to Isle Royale in 1948, moose 
numbers have remained between 500 and 1 ,600 
(Krefting 1974, Peterson 1992) (Figure 3). Wolf 
numbers remained in the low twenties through the 
1960s, then increased in the 1970s (to 50 in 1980) 
and declined in the 1980s (to 12 in 1989) (Peterson 
1990). As of March 1992, moose numbered about 
1 ,600 and wolves still numbered 12 (Peterson 
1992). 



Background 



I 



60 



50 



[2 40 

> 

_i 

§ 30 



20 



10 




2000 
1800 
1600 



1 



1400 uj 

CO 

1200 O 
1000 5 
800 
600 



59 64 69 74 79 84 89 94 

YEAR 



400 



Figure 3. Moose and wolf populations on Isle Roy ale since 1959 (from Peterson 1992). 



Isle Royale National Park was established in 1931; 
since 1977, 99% of the land area has been formally 
designated as wilderness (Karamanski et al. 1988). 
Humans have used Isle Royale, however, for over 
4,500 years (Crane and Griffin 1965). Archeologi- 
cal evidence of seasonal hunting, fishing, maple 
sugaring, and copper mining during the Archaic and 
Woodland periods has been found at numerous Isle 
Royale sites (C. Clark, pers. comm.). 

In historic times, hunting and trapping for subsis- 
tence and for the fur trade, as well as a commercial 
fishing operation for the American Fur Company, 
preceded the Ojibwa Indians' cession of Isle Royale 
to the United States in 1842 (Karamanski et al. 
1988). Subsistence use by north shore Lake Supe- 
rior Ojibwas continued at least into the 1870s. After 
cession, Euro-American settlement came haltingly 
with three copper exploration booms (in the 1 850s, 



early 1870s, and early 1890s), Scandinavian 
fishermen, and finally tourism and summer homes 
in this century (ibid). 

Presently, Isle Royale has no roads or year-round 
residences. Dwellings are limited to park headquart- 
ers on Mott Island, two main visitor entry areas at 
Rock Harbor and Windigo, five outlying ranger 
stations, and about 16 active life leases to summer 
residents, principally in Tobin Harbor (Figure 4). 
Approximately 1 5,000 persons visit Isle Royale 
each year from May to October, either staying at 
Rock Harbor Lodge, in private boats, or hiking and 
camping on the 267-km trail system. Camping is 
restricted principally to designated campgrounds, 
including 1 2 inland sites, 1 8 shoreline sites, and 6 
sites on offshore islets. Winter occupancy is limited 
to the wolf-moose research team from mid-January 
to early March. 



Introduction 



ISLE ROYALE NATIONAL PARK 



^ 



Anygd< 



Canpsite 
Trail 




Ma lone Bay 



50 km 



Figure 4. Isle Royale National Park map. Map: Eric Gdula. 



Caribou Distribution and Decline 
in the Lake Superior Region 

Before the fur trade and settlement caused major 
shifts in large mammal distributions in the 19th 
century, woodland caribou ranged southward from 
Lake Superior to central Minnesota and Wisconsin, 
lower Michigan, and southern Ontario (Figure 5) 
(Bergerud 1978, Baker 1983, Darby et al. 1989). 
During the late 1800s and early 1900s, the caribou 
range retreated steadily northward until these large 
ungulates disappeared from northeastern Minnesota 
and Isle Royale in the 1920s, and from northwestern 
Minnesota and Ontario north of Lake Superior to 
Lake Nipigon in the 1940s and 1950s. Relict herds 
have persisted in Ontario south of 50° latitude only 
on the Slate Islands, Pic Island, Pukaskwa National 
Park on the north shore of Lake Superior, and a few 



inland sites (Darby et al. 1989) (Figure 6). To- 
gether, these herds numbered approximately 500 
animals in 1990 (Abraham et al. 1990). Recently 
caribou have been successfully translocated from 
the Slate Islands to Michipicoten Island and to 
Leach and Montreal islands and the Gargantua 
Peninsula in Lake Superior Provincial Park farther 
east on the Lake Superior north shore (Figure 6; 
Appendix C) (G. Eason, pers. comm.). 

The woodland caribou's decline in eastern North 
America has not been explained conclusively. 
Bergerud (1974a, 1978; Bergerud and Mercer 
1 989) presents the following hypothesis for the 
northward range retraction of caribou. Woodland 
caribou have a comparatively low reproductive rate 
for an ungulate, which balances closely with natural 
mortality in undisturbed populations. When Euro- 



Background 



1990 




pre-1800 



■ Relict Herd 
▲ Introduced Herd 



Figure 5. P resettlement and current range limits of woodland caribou in the Great Lakes region. Sources: 
Bergerud 1978, Baker 1983, Darby et al. 1989. 



Americans brought firearms to native hunters and 
increased overall hunting pressure, caribou began to 
decline. At the same time, extensive forest clearing 
and burning were greatly expanding early succes- 
sional habitats. This trend, possibly combined with 
a period of climate warming after 1860, facilitated 
the northward range expansion of white-tailed deer 
and the increase of moose and deer populations. 

As the expanding numbers of ungulates increased 
total available prey biomass, predators, especially 
wolves, responded by increasing. Thus, both 
human-caused and natural mortality increased, 
perhaps sequentially, beyond the capacity for 
caribou to replace their losses. Bergerud ( 1 978) 
explains that hunting mortality is typically additive 
to natural mortality in caribou. Thus, often "there is 
no surplus for hunters and any hunting will reduce 
future populations. In addition, population declines 
are difficult to halt. . . . Hunters can still find 
caribou [when they are rare]" (Bergerud 1978:93- 
94). White-tailed deer also carried a parasitic 



disease highly lethal to caribou, the meningeal 
brainworm. This parasite has eliminated caribou 
from any significant range overlap with white-tailed 
deer (Bergerud 1978). 

Applying this hypothesis to the region south of Lake 
Superior, caribou would have been first severely 
overhunted and then exposed to the meningeal 
brainworm before the late 1 800s. On the north 
shore of Lake Superior, overhunting was severe as 
early as the late 1700s (Lytwyn 1986). White-tailed 
deer invaded the fringes of this region after 1900, 
but high moose numbers, and hence, predator 
densities, are likely the chief control on caribou 
populations in northwestern Ontario at present 
(Abraham et al. 1990). Relict herds have persisted 
south of 50° latitude only where deer are few and 
refuge habitat insulates caribou from wolf predation 
(Bergeaid 1974c/). 



Introduction 














40 80 




\p^v^N^£ic lsland 

" Slate ^ 

Islands ^| iPukaskwa 




km 






fs 


Iv National Park 
Michipicoten j^ m^m 






Isle Royale 


Island flPLake Superior 

^_^~- >. ^WProvincial Park 
f^ Montreal »J 






LAKE / 


/ Island C 
J SUPERIOR C^.^ 








^— "~**- P/ 



Figure 6. Relict and introduced woodland caribou around Lake Superior. Source: Darby et al. 1989. 



Caribou on Isle Royale 

Woodland caribou were present on Isle Royale for 
more than 3,500 years until they disappeared in the 
late 1 920s. Their decline paralleled the species' 
retreat from northeastern Minnesota and Ontario 
north of Lake Superior. The caribou's decline on the 
island coincided with their disappearance from the 
nearest mainland, Ontario's Sibley Peninsula (de 
Vos and Peterson 1951, Cringan 1 956), and these 
events are certainly related. However, unlike 
conditions on the mainland, the immediate cause of 
the Isle Royale caribou extirpation could not have 
been meningeal disease or wolf predation. Wolves 
were absent and white-tailed deer — although 
introduced in 1 9 1 — were few or absent on Isle 
Royale in the 1920s. 

Caribou were probably overhunted on Isle Royale 
during the 1 800s. Native Americans had a tradition 
of hunting on Isle Royale, which became an even 
more valuable resource when mainland game 
resources were nearly exhausted in the early 1800s 
due to the fur trade. Subsistence hunting by Euro- 



American miners and fishermen likely contributed 
to low caribou numbers on Isle Royale in the mid- 
tolate 1800s. 

After 1900, regional faunal changes spilled over 
onto Isle Royale, bringing a new competitor, moose, 
and a new predator, coyote, and isolating the Isle 
Royale caribou herd from mainland populations. 
Forest burning from 1 850 to 1 890 by mineral 
prospectors facilitated a population irruption of 
immigrant moose. By the late 1920s, moose over- 
browsing could have limited caribou food supplies. 
But caribou were already near extirpation by that 
time, suffering from coyote predation in addition to 
lynx predation, and lacking traditional immigrants 
from the mainland to replenish their reduced 
numbers. Although a few white-tailed deer were 
introduced to Isle Royale in 191 0, they never 
prospered and were unlikely to have transmitted the 
meningeal brainworm to caribou. However, caribou 
traveling off Isle Royale to the mainland could have 
been exposed to the parasite. 



Background 



The density of wolves on Isle Royale is still high 
compared to most of their mainland range — 12 
wolves on 544 km 2 , or more than twice the maxi- 
mum density that Bergerud and Mercer ( 1 989) 
predicted introduced caribou could tolerate. How- 
ever, caribou escape habitat is available in an 
archipelago of offshore islands. Thus, Bergerud 
(pers. comm.) predicted that wolves might continue 
to depend on comparatively abundant moose, 
ameliorating predation pressure on caribou if they 
were restored to the island. Further, wolf research- 
ers predict that wolf numbers are likely to continue 
a long-term downward trend (R.O. Peterson, pers. 
comm.). This decline in wolf numbers led Bergerud 
and Mercer ( 1 989: 1 1 8) to conclude that "now 
would be an excellent opportunity to reintroduce 
caribou to this National Park." 



Introduction 



Methods 



Following is a summary of the methods used for 
this study and an overview of the Result s and 
Discussion section. 

Why Did Caribou Disappear from Isle Royale? I 
drew many historical references from Martin's 
(1988) draft report on the history and restoration 
potential of three Isle Royale extirpated species, 
marten, lynx, and caribou. Further research led to 
numerous sources in the Isle Royale National Park 
archives, unpublished fur trade records, the Michi- 
gan Historical Collections at the University of 
Michigan's Bentley Historical Library, and a 
variety of published literature. 

Habitat Suitability. The initial proposal for this 
feasibility study called for adapting a northern 
Minnesota woodland caribou habitat suitability 
index model (HSI) to Isle Royale and assessing Isle 
Royale caribou habitat through the model. How- 
ever, the model was not completed as expected in 
1990. Thus, I completed a preliminary assessment 
of habitat characteristics for this report, drawing on 
published sources, park maps, some field reconnais- 
sance, and consultation with caribou biologists. In 
the future, quantitative habitat evaluation will be 
feasible using the park's new geographic informa- 
tion system (GIS). 

Population Vulnerability Analysis. I developed a 
synopsis of typical and projected demographic 
characteristics for a hypothetical caribou herd, 
based on published literature and advice from 
caribou biologists. Then, with assistance from Dr. 
Terry Kreeger of the University of Minnesota, I 
employed a computerized population model ("VOR- 
TEX," developed by Robert Lacy of the Chicago 
Zoological Park, Brookfield, Illinois; see Appendix 
B) to project the long-term viability of this hypo- 
thetical population. Repeated simulations resulted in 
a mean persistence time, probability of persistence 
to 100 years, and retention of genetic heterozygosity 
for different initial release alternatives and mortality 
rate projections. 



Re establishment Methods. I reviewed alternative 
reestablishment approaches from preceding caribou 
translocation projects in Canada and the United 
States. These methods are outlined only briefly in 
this report, based on the advice of experienced 
biologists and an understanding of the logistical 
constraints of work on a wilderness island. Detailed 
procedures would have to be developed for a 
restoration plan. 

Impacts and Management Implications of Caribou 
Reestablishment. I tentatively identified the likely 
impacts of caribou releases on park vegetation and 
other animals through a review of pertinent litera- 
ture, communication with other biologists, and 
familiarity with park resources. I discussed the 
relationship of this project to other resource man- 
agement activities with park staff. I also considered 
examples from other parks, such as Pukaskwa 
National Park in Ontario. This report provides only 
a preliminary overview of potential impacts and 
management concerns that would have to be 
addressed in much greater detail if caribou restora- 
tion is subsequently pursued by the National Park 
Service. 



Results and Discussion 



Why Did Caribou 
Disappear From Isle 
Royale? 

In this section, I outline and critique six hypotheses 
for why caribou disappeared from the park: migra- 
tion, disease, competition with moose, predation, 
overhunting, and fires. This analysis is based on the 
detailed history of Isle Royale caribou presented in 
Appendix C. 



immigration may even have been essential to the 
long-term persistence of Isle Royale's relatively 
small herd of at most a few hundred animals. 
Immigrants would have prevented inbreeding 
depression and replenished the herd after "bottle- 
necks" or catastrophic declines, which are common 
in small populations (Gilpin and Soule 1986). Thus, 
Isle Royale supported a resident caribou herd, but 
over time this herd may have been dependent on a 
flow of animals from the mainland. Severing this 
flow could have contributed to the demise of the 
island's remnant herd. 



Migration 

The coincident timing of the caribou's disappear- 
ance from Isle Royale with the decline of caribou 
herds on the adjacent mainland has fostered a theory 
that these mobile animals may have been seasonal 
migrants to the island archipelago (Stoll \924b, 
Mech 1966, Martin 1988). Yet caribou were clearly 
year-round residents on Isle Royale 
because they were documented there 
during summer as well as winter and 
spring. Lake Superior freezes infre- 
quently between Isle Royale and the 
mainland and regular migration by 
swimming such a wide channel would 
be unlikely. Further, the environment 
on Isle Royale was probably suitable 
for caribou, providing no apparent 
impetus for a risky migration to 
potentially less favorable calving 
grounds on the mainland. 



Disease 

Martin (1988) suggested that migrants may have 
carried meningeal brainworm or other diseases to 
Isle Royale caribou, but we have no evidence for 
this hypothesis. The gradual decline of the island's 
caribou does not describe a population suddenly 



Yet the Isle Royale herd was not 
isolated. Caribou are well adapted for 
travel on ice and they were observed 
on the lake ice between Isle Royale 
and both Minnesota and Ontario 
(Figure 7; Appendix C). Periodic 




Figure 7. Woodland caribou crossing a frozen lake. Woodland 
caribou will cross large lakes over the ice, and a number of historical 
accounts indicate that caribou crossed the area between Isle Royale 
and the adjacent Ontario mainland. Photograph: H.R. Timmermann, 
Ontario Ministry of Natural Resources. 



11 



exposed to a fatal parasite. White-tailed deer have 
not immigrated to Isle Royale, but they were 
introduced in 1910 by a Michigan state game 
warden (Wood and Dice 1924 in Martin 1988). 
Summer resident Frank Warren (1926, 1929) 
reported that seven or nine animals were released 
and individuals or tracks were sighted as late as 
1925. The Michigan State Game Commission's 
reports that the herd "increased wonderfully" 
{Detroit News editorial, 3 December 1921) were 
refuted by others. For example, island visitor and 
journalist W.P.F. Ferguson (1922a) wrote in a letter 
to Albert Stoll, Jr.: "The [game commission's] 
estimate concerning the deer on the island is errone- 
ous. Lively, the game warden, tells me that he has 
seen only two or three. They cannot live with the 
[brush] wolves [meaning coyotes]." 



initial irruption of the moose herd on Isle Royale, is 
that moose overbrowsing eliminated winter food for 
caribou (Peterson 1977, Martin 1988). Krefting 
(1974) summarized the history of moose on Isle 
Royale. Moose arrived by swimming or crossing the 
ice around 1905, possibly earlier, having become 
common on the north shore by the 1 890s. A solid 
ice bridge in the winter of 1912-13 was a likely 
source of more immigrants. By 1915, the population 
was well established at 250-300 animals and 
numbers increased steadily thereafter. Krefting 
(1974) believed the best estimates for the moose 
population were 1,000 in 1921-22, 2,000 in 1925- 
26, and from 2,000 to 5,000 by 1930. The inevi- 
table die-off began in 1933; moose numbers bot- 
tomed out at a reported 200 by 1935 (Krefting 
1974). 



In 1916-17, W.H. Foster (1917), a warden, estimat- 
ed that only 20 deer persisted on Isle Royale. 
Warren (1929) wrote to Albert Stoll, Jr., that 
coyotes had killed all the deer by 1929, a claim Dr. 
Frank Oastler (1929) supported after his survey of 
Isle Royale for the National Park Service. However, 
residents such as fisherman Ed Holte (Holte and 
Holte 1965) recalled seeing deer at a salt lick at 
Pickerel Cove in the early 1930s and that three deer 
were on Amygdaloid Island during the 1936 fire. 

Deer were not reported after 1936 and never 
approached the density of 0.4/km 2 reportedly 
required for meningeal brainworm transmission to 
caribou populations (Karns and Lindquist 1986). 
Also, brainworm disease symptoms have never been 
reported in Isle Royale moose, including during the 
1930s moose die-off. A report on meningeal brain- 
worm larvae collected from Isle Royale moose feces 
(Karns and Jordan 1969) was subsequently refuted 
as an error (Lankester and Hauta 1989). 



Competition with Moose 

Another possible explanation for the caribou's 
disappearance from Isle Royale, based on the 
coincident timing of caribou extirpation with the 



The impact of moose browsing on the food supplies 
of Isle Royale caribou is difficult to surmise. In 
1917, warden Foster (1917) reported that "lichens, 
which festoon the swamp trees profusely, are. . . [the 
moose's] principal diet." He said the caribou were 
also "lichen and moss eaters," and "edible lichens 
and tree mosses are abundant." When refuting the 
high game animal estimates published in the Detroit 
News, Frank Warren (1924) wrote: 

It is my best judgement and belief that the numbers 
of moose and caribou have been grossly overstated 
[e.g., by Albert Stoll, Jr.]. It will be many years be- 
fore they crowd each other for feed (largely browse) 
except in their favorite places. There are many thou- 
sands of acres which I have seen where there are 
hardly any moose sign, where young poplars and 
birch abound. 



At the time the last caribou were sighted on Isle 
Royale in 1928, moose density may have reached 4/ 
km 2 . By 1930, Murie (1934) recorded overbrowsing 
on all of the important moose winter foods, such as 
balsam fir, quaking aspen, paper birch, and Ameri- 
can mountain ash (Sorbus americana). However, 
moose and their impacts were concentrated toward 
the west end of Isle Royale (Krefting 1974). While 
living on the east end of Isle Royale, Warren (1929) 



12 



Results and Discussion 



did not see signs of moose browse on balsam fir 
until 1927 and in 1929, claimed overbrowsing was 
still localized. 

By this time, Canada yew (Taxus canadensis), a 
primary winter food for caribou as well as moose, 
had been reduced from understory dominance to 
virtual absence on the main island (Murie 1934, 
Janke et al. 1978). Yet yew was still abundant on 
offshore islands, including Mott Island and Wright 
Island, as late as 1931 (Krefting 1974). A 1930 
photograph of the "Desor Trail" in Albert Stoll, 
Jr.'s, collection shows a forest festooned with 
arboreal lichens {Alectoria and Usnea spp.). The 
condition of other caribou winter browse, such as 
ericaceous shrubs and ground lichens (Cladonia 
spp.), was not documented. 

In summary, while the absolute abundance of 
caribou foods was diminished by the late 1920s, 
starvation is still an improbable explanation for the 
caribou's demise on Isle Royale. Competition with 
moose cannot have been the sole nemesis of Isle 
Royale's caribou because 

1 . caribou began to decline before moose were 
abundant (numbering as few as 30 when 
moose numbered only 250-300); 

2. moose do not eat all of the plants that 
caribou can consume (e.g., ground lichens); 

3. prime winter caribou browse was still 
available on offshore islands and, to a lesser 
extent, inland when caribou disappeared; 

4. browse was still sufficient for moose num- 
bers to increase or be sustained for about 
five years after caribou disappeared; 

5. caribou are highly resilient to winter food 
shortages (A.T. Bergerud, pers. comm.). 

Corroboration for this conclusion comes from the 
Slate Islands where caribou have survived for 
decades on minimal winter browse supplies — 
principally windthrown arboreal lichens (A.T. 



Bergerud, pers. comm.). Also, caribou can maintain 
normal fertility rates despite range depletion (Berg- 
erud 1980, 1983). Reindeer introduced to islands 
have increased to densities over 12/km 2 before 
starving, without an appreciable reduction in annual 
productivity (Klein 1968). 



Predatlon 

Moose and disease were not the only potential 
threats arriving from the mainland in the period of 
great regional faunal change after 1900. Coyotes 
also arrived on Isle Royale, at least by 1912 and 
possibly by 1906 (Krefting 1969). By 1916-17, 
trappers were able to capture 60 "brush wolves," 
leaving 12 or more (Foster 1917). Krefting (1969) 
cited reports of increasing coyotes from 1918 to 
1925, despite trapping by fishermen and state game 
wardens. 

Coyotes can be effective predators on caribou 
calves. For instance, newly immigrated coyotes, in 
conjunction with black bears (Ursus americanus), 
have been implicated for the recent population 
decline of the isolated Gaspe Peninsula caribou herd 
in Quebec (E. Mercer, pers. comm.). 

Isle Royale residents blamed coyotes for white- 
tailed deer failing to survive and even the dearth of 
moose calves in the park in the late 1920s (Fergus- 
on 1922c, Warren 1929). Michigan Conservation 
Department official Hugh E. Green visited Isle 
Royale in 1928 "in the interest of improving game 
conditions in Isle Royale, as it had been reported 
that coyotes are becoming so numerous they 
threaten other wildlife" (Anonymous 1928). Preda- 
tor control programs were common in this era and 
were apparent by the series of game wardens 
(trappers) that were placed on Isle Royale after 
1916. 

At the same time coyotes were colonizing Isle 
Royale, lynx were still commonly seen (Martin 
1988). For example, Foster (1917) reported 67 lynx 
captured in 1916-17, and fisherman such as Milford 



Why Did Caribou Disappear From Isle Royale? 



13 



Johnson (1965) recalled that lynx 
were common and trapped occasion- 
ally in the decades before 1930. 
During periodic snowshoe hare 
population crashes, Isle Royale lynx 
may have been heavily reliant on 
caribou calves for alternate prey. 
Bergerud ( 1 97 1 ) has demonstrated 
that lynx can control caribou herd 
recruitment during these phases. 

Fisherman Pete Edisen (Edisen and 

Edisen 1965) raised the possibility of 

a third caribou predator. Pete recalled 

that most of the island residents 

owned dogs, an observation borne out 

in numerous photographs. These dogs 

bred with the "brush wolves" and ran 

in packs (ibid). Feral dogs can be 

efficient hunters. If they were roaming Isle Royale 

as Pete suggested, they could have contributed to 

the caribou herd's predation problems. 



Overhunting 

None of the 20th-century invaders mentioned 
earlier — the hypothetical competitor, disease, or 
predators — explains why caribou numbers were 
apparently so low on Isle Royale in the 19th centu- 
ry. In North American boreal forests, caribou 
densities average 0.3-0.4/km 2 in association with 
natural predator numbers and scarce alternate prey 
(Bergerud 1980, 1983, pers. comm.). Covering 544 
km 2 , providing ample food supplies, and having 
only lynx for a predator, Isle Royale should have 
supported at least an average density of caribou in 
the 19th century— or 200 animals. Stoll's (1926) 
population estimates for pre- 1920, 200-400 ani- 
mals, exceeded this average boreal density. Other 
observers may have underestimated the number of 
these reclusive animals on Isle Royale, a common 
bias for estimates of dispersed animals in forested 
habitats (Hickie n.d.). In contrast, the evidence 
points to low caribou numbers on Isle Royale before 
1900, specifically 1840-90. 




Figure 8. Caribou hunters on the north shore of Lake Superior in 
1925 or 1926. Photograph: O. Anderson, courtesy ofH.R. 
Timmermann, Ontario Ministry of Natural Resources. 



These 19th-century dates coincide with a period of 
overhunting in the Lake Superior region, from the 
end of the fur trade through the peak of Euro- 
American settlement. Native American hunters 
would certainly have been drawn to Isle Royale 
when game was depleted elsewhere by the 1830s, 
especially given their tradition of visiting the island 
and Isle Royale's reputation for abundant game. 
Later in the 19th century, miners and fishermen 
would have been both well armed and typically 
short of red meat, with plenty of time in the winter 
to shoot caribou (Figure 8). Thus, despite finding 
few documented records, I conclude that caribou 
were probably hunted regularly on Isle Royale 
before 1900. Elsewhere, annual hunting mortality as 
low as 5% of a population has initiated caribou 
population declines (Bergerud 1980). 



Fires 

The extensive forest fires ignited by mineral explor- 
ers periodically between 1850 and the 1890s altered 
forest composition and successional stages on Isle 
Royale (Hansen et al. 1973, Janke et al. 1978). By 
reducing lichen biomass, fires can reduce habitat 
suitability for caribou for many years (Bergerud 
1978, Abraham et al. 1990, Schaeffer and Pruitt 



14 



Results and Discussion 



1991). In the long run, fires in mature conifer 
forests should benefit caribou by restoring ground 
lichen biomass (ibid). In habitats where caribou are 
not dependent on lichens for winter browse, the 
negative effects of fire may be minimal. Thus, fire 
may be a partial but secondary explanation, along 
with overhunting, for the caribou's decline on Isle 
Royale after 1840. 



Summary 

In terms of conservation biology, overhunting and 
perhaps fires in the 19th century most likely began 
an "extinction vortex" (Gilpin and Soule 1986) for 
Isle Royale's caribou by reducing their numbers 
below a critical minimum level. The subsequent 
disappearance of caribou on the adjoining mainland 
shores broke a link to continental populations, 
assuring this remnant herd's isolation. An isolated 
herd of less than 50 animals could have vanished 
over time merely by chance demographic events, 
such as a series of years with poor recruitment, or 
from a random catastrophe, such as most of the 
animals wandering off toward Canada on thin ice. 

I speculate that if numbers were actually as low as 
30, as game warden Foster (1917) estimated in 
1917, then the most likely cause of the caribou's 
final disappearance from Isle Royale was coyote 
and lynx predation. However, if numbers were still 
as high as 200-300 just before 1920, as Stoll (1926) 
suggested, then caribou were more likely to have 
suffered from direct competition with moose in 
addition to increasing predation. In any case, 
reduced food supplies could have exacerbated the 
downward population vortex by reducing individual 
fitness. 

In conclusion, humans contributed to extirpating 
Isle Royale's caribou both directly and indirectly. 
Native American and Euro- American hunting 
reduced caribou numbers, perhaps greatly, in the 
19th century. Extensive, human-caused forest fires 
may also have reduced Isle Royale's carrying 
capacity for caribou for some time. Overhunting 



and forest clearing on the mainland removed 
caribou from the adjacent shore and isolated the Isle 
Royale herd. By creating early successional habitats 
on the mainland, settlers also facilitated moose and 
coyote immigration to Isle Royale. Once on Isle 
Royale, moose and coyotes benefited from the 
expanse of early successional habitat created by 
miner's fires, at the expense of caribou. Thus, 
numerous, seemingly "natural" events — range 
expansions, disease, and predator responses — were 
played out on a landscape that human activity had 
greatly altered. 



Habitat Suitability 

The habitat suitability index model for woodland 
caribou in northern Minnesota, being prepared 
under contract to Voyageurs National Park (Gogan 
et al. 1991), was not completed in time for this 
study as anticipated. Hence, the following discus- 
sion is an overview of habitat resources on Isle 
Royale based on (1) prior vegetation surveys and 
maps (Hansen et al. 1973, Slavik and Janke 1987); 
(2) measurements of offshore islets using the 
nascent Isle Royale GIS at Michigan Technological 
University; (3) site visits on the northeast end of Isle 
Royale; and (4) assessments by biologists familiar 
with caribou, Isle Royale, or both. 



Forage Resources 

Slavik and Janke (1987:94-95) provided an over- 
view of Isle Royale plant communities, as follows: 

Isle Royale is densely forested. Two distinct forest 
types reach their climax there. The island is situ- 
ated at the southernmost limit of the northeastern 
coniferous forest. The spruce/fir/paper birch boreal 
forest type . . . achieves its climax here at the lower 
elevations around the periphery of the island where 
climatic conditions are moister and cooler (Linn 
1957). . . . Other tree species commonly encountered 
in the boreal forest are northern white cedar and 
quaking aspen. Common ground cover plants found 



Habitat Suitability 



15 



in the mature northern boreal forest are large-leaved 
aster (Aster macrophyllus), wild sarsaparilla (Ara- 
lia nudicaulis), Canada dogwood (Cornus cana- 
densis), clintonia (Clintonia borealis), twinflower 
(Linnaea borealis) and bristly club moss (Lycopo- 
dium annotinum). Tall shrubs are represented by 
thimbleberry (Rubus parviflorus) and mountain al- 
der (Alnus crispa). 



The second climax association is found on the west- 
ernmost end of the island. At higher elevations away 
from Lake Superior, where the soil is drier and the 
air warmer, the sugar maple and yellow birch asso- 
ciation is the climax forest type (Linn 1957). Sugar 
maple is at its northernmost limit here as part of 
the temperate deciduous forest biome. This smaller 
climax association covers only about 7% of the is- 
land. . . . These two upland forest types which domi- 
nate Isle Royale are in sharp contrast to the low- 
lands between the parallel system of ridges. These 
contain a vast amount of wooded swamp composed 
of black spruce and northern white cedar. 



Also in these lowland areas many bogs are found. 
The bog communities of Isle Royale are rich in plant 
life, with sedges often dominating the sphagnum 
mat. . . . 



In 1936 20% of the island was burned and the hu- 
mus layer was destroyed. This region, stretching 
from Siskiwit Bay in the west to Moskey Basin in 
the east, is now almost a pure stand of paper birch 
with scattered quaking aspen. The abundant ground 
cover species here are large-leaved aster, thimble- 
berry, bracken fern (Pteridium aquilinum), and wild 
sarsaparilla. 



In the late 1960s, Krefting et al. (1970) produced a 
forest cover-type map of Isle Royale, which is 
described in greater detail by Hansen et al. (1973) 
(Figure 9). They recognized 27 vegetation types in 7 
cover groups on Isle Royale (Table 1). 




N 

1 



Vegetation Type 

llll white spruce, balsam fir 
E3 yellow birch, sugar maple 
IS 1936 fire (birch, aspen) 
LLl jack pine, black spruce 
I I lakes 



10 



km 



Figure 9. Vegetation type map of Isle Royale National Park. 



16 



Results and Discussion 



Table 1. Isle Royale vegetation and nonvegetation cover types and areas covered. 



Cover Type Description 


Hectares 


Percent 




Covered 


Covered 


Quaking aspen, paper birch, balsam fir, white spruce 


27,731 


51 


1936 burn 


10,505 


19 


Black spruce, northern white cedar (and balsam fir, tamarack 


5,570 


10 


[Larix laricina], or both) 






Yellow birch, sugar maple (and paper birch, balsam fir, or both) 


4,161 


7 


Lakes (30 named) 


3,415 


6 


Rock outcrop 


2,011 


4 


Jack pine (and black spruce) 


177 


1 


Shrubs 


503 


1 


Beaver ponds 


477 


1 


Total 


54,550 


100 



Adapted from Hansen et al. 1973. 



Moose browsing has profoundly affected forest 
composition on Isle Royale (Snyder and Janke 
1976, Krefting 1974). For example, moose had 
virtually eliminated Canada yew from the main 
island by 1930, where it had been the dominant 
shrub (Murie 1934, Snyder and Janke 1974). After 
more than 60 years of heavy browsing, balsam fir, 
mountain ash (Sorbus americana), and shrubs such 
as red osier dogwood (Cornus stolonifera), squash- 
berry (Viburnum edule), and highbush cranberry 
(Viburnum trilobum) have been reduced in domi- 
nance in the boreal forest type (Snyder and Janke 
1974, Jordan 1978). Only white spruce is increasing 
among the main tree species, because it is not eaten 
by moose. 

The net effect of moose browsing is a more open 
forest, with more abundant ground cover due to 
increased light penetration (Snyder and Janke 
1974). Forage available to moose also declines 
unless replenished by fire or another disturbance 
(Figure 10). However, bryophyte and lichen cover 
are apparently unaffected by moose browsing (ibid). 
Despite the substantial effects moose have had on 
the Isle Royale forest, browse supplies have been 
sufficient to support a large moose herd for decades. 
Moose density has averaged 2.9/km 2 (Peterson 



1992). Undercurrent vegetation trends, carrying 
capacity for moose should decline gradually — but 
unpredictable events could alter this trend. 

Snyder and Janke (1974) demonstrated that the 
effects of moose browsing have been considerably 
less on offshore islets than on the main island, 
especially on islets that are less frequently connect- 
ed to the main island by ice. Thus, more remote 
islets remain barely touched by moose (Figure 11). 

Woodland caribou and moose diets overlap some- 
what. Yet, caribou usually bypass deciduous twigs 
and balsam fir, which are the staples of moose 
winter diets (Cringan 1957, Simkin 1965, Bergerud 
1972, Darby and Pruitt 1984, Edmonds and Bloom- 
field 1984). Alternately, plants rarely taken by 
moose form the core winter diet of caribou — 
arboreal and terrestrial lichens, sedges, and ericoids. 
On Isle Royale, these typical caribou winter foods 
are not regularly eaten by moose, although moose 
do eat some arboreal lichens in the park (R.O. 
Peterson, pers. comm.). Bogs are scattered all 
across the main island and surrounding islets, 
providing patchy supplies of bog ericoids and 
sedges (Stardom 1975, Brown and Theberge 1990) 
(Figure 12). 



Habitat Suitability 



17 




Figure 10. Savanna habitat near Windigo created by intensive 
moose browse. Moose have virtually eliminated the shrub layer in 
some favored habitats on Isle Roy ale. Photograph: R. Janke, Isle 
Royale National Park archives. 




Figure 11. Lightly browsed habitat on Smithwick Island. In the early 
1970s, moose had not browsed the vegetation on Smithwick Island. Yet 
by 1990, when this photograph was taken, favored species such as yew 
and mountain ash showed signs of browsing, indicating that moose 
had wintered on this Rock Harbor island in recent years. Still, forage 
supplies remain abundant on Smithwick Island and other islets that 
moose do not regularly occupy. Photograph by the author. 



18 



Results and Discussion 





vni&Lfc 






W£*t^« 



raK^n^^^H 






Figure 12. Bog on Raspberry Island. This 
bog is typical of Isle Roy ale's numerous, small 
black spruce bogs. Photograph by the author. 




Figure 13. Jack pine stand with ground lichens, Saginaw Peninsula. 
Isle Royale's only substantial stand of jack pine is on the Saginaw 
Peninsula. This forest has not burned since the 19th century. Ground 
lichens are a common ground cover in this forest type. Photograph by 
the author. 



Habitat Suitability 



19 



Ground lichen supplies maybe more sparse on Isle 
Royale than in northern Ontario. Although widely 
scattered in rocky openings throughout Isle Royale 
and especially along south-sloping shorelines, the 
only extensive ground lichen mats are in the jack 
pine and black spruce stands on the Saginaw 
Peninsula (R.O. Peterson, pers. comm.) (Figure 13). 
In the late 1960s, jack pine forest stands covered 
only 177 ha (0.32%) of Isle Royale (Hansen et al. 
1973). Additional, small outcrops supporting jack 
pine and ground lichen mats that were not mapped 
by Krefting et al. (1970) are scattered around the 
east end of Isle Royale. The Saginaw Point jack 
pine stands have not burned since the 1 890s so 
ground lichen biomass may be diminished from 
historical levels. Subsequent burns in this fire- 
adapted community could eventually improve 
forage conditions for caribou by restoring the 
ground lichen cover (Bergerud 1978, Abraham et al. 
1990, Schaeffer and Pruitt 1991). 

Arboreal lichens are common in the boreal forest 
type, especially in the cool, moist microhabitats 
near the lakeshore and on islets (Figure 14). Arbo- 
real lichens are most noticeable in mature balsam 



and spruce trees and less common in the decadent 
birch and aspen stands or densely stocked young fir 
stands. 

Summer foods for caribou are abundant on Isle 
Royale. In particular, thimbleberry, which has 
become the dominant understory shrub across much 
of Isle Royale because moose avoid eating it, would 
provide summer browse for caribou (A.T. Bergerud, 
pers. comm.). Similarly, caribou could take advan- 
tage of the expansive quantities of large-leaved 
aster, bracken fern, Canada dogwood, twinflower 
(Linnaea borealis), and numerous other herbs as 
well as mosses and lichens that are not favored by 
moose. Even shrubs and saplings that are eaten 
readily by moose, such as the viburnums, Cornus 
spp., mountain maple, mountain ash, and red- 
berried elder (Sambucus pubens), are still present 
parkwide, especially on islets (Figure 14). 

In general, caribou are able to survive in mature 
forests better than moose, which thrive on the 
earliest forest successional stages (Abraham et al. 
1990). While caribou are not obligate climax forest 
animals, they are able to exploit foods such as 
arboreal lichens that are typical of mature, northern 
coniferous forests. 




Figure 14. Spruce-fir forest on Mott Island. The white spruce and 
balsam fir forest on Mott Island is typical of the eastern end of Isle 
Royale. Large-leaved aster is prominent where the understory is partly 
shaded, while thimbleberry grows best in direct sunlight. Photograph 
by the author. 



Much of Isle Royale was severely 
burned during the 19th century 
(Rakestraw 1965). While these forests 
have not yet reached climax stages, 
they are now 100 years old or older 
(Snyder and Janke 1974). Trees 
regenerating after the 1936 fire grew 
out of the reach of moose in the 
1960s, and the 1936 burn area has 
supported their lowest densities on 
Isle Royale since then (R.O. Peterson, 
pers. comm.). Similarly, moose 
bypass the smaller 1948 burn area 
where minimal browse is available. 
Herbaceous, summer forage may be 
adequate for a low density of caribou 
in these 45-60-year-old birch and 
aspen stands that cover over 20% of 
the park. 



20 



Results and Discussion 



Predators 

Predator densities are a key predictor 
of habitat suitability for caribou 
(Bergerud and Elliot 1986, Bergerud 
and Mercer 1989). On Isle Royale, the 
predators are wolves and potentially 
lynx. Coyotes and black bears are 
absent. 

To generate a "worst case scenario" 
for caribou restoration, I assumed that 
wolves would be reestablished if 
current trends continue and they 
become extirpated from Isle Royale. 
In fact, the National Park Service will 
not decide whether to reestablish 
wolves until the reasons for their 
extirpation, presuming they disappear, 
are understood and a feasibility study 
and environmental assessment for wolf restoration 
are completed (R.J. Krumenaker, pers. coram.). If 
wolves do not recover on their own and are not 
reestablished, the predictions in this report for 
caribou restoration, based on high predator num- 
bers, would have to be revised. 

Wolves arrived on Isle Royale in the late 1940s 
(Figure 15) and have numbered between 12 and 50 
animals (20-92/1,000 km 2 ) since 1950 (Peterson 
1992, Figure 3). Wolf numbers peaked at 50 in 
1980 after a decade of preying on abundant moose. 
Then suddenly the wolf population crashed by more 
than 70% in two years. Subsequent research has 
identified starvation, interpack fighting due to food 
shortages, and possibly canine parvovirus as the 
likely causes (Peterson and Thurber 1990). By the 
mid-1980s wolves had increased again. But in 
1985, wolves began another steady decline due to 
high annual mortality and declining reproduction. 
Intensive research on the causes of the wolf decline, 
including radio-collaring and blood assays of 
wolves, began in 1988. 

Isle Royale wolf research has not yet ruled out the 
possibility that food shortage is responsible for the 
low wolf reproduction rates in the 1 980s (Peterson 




Figure 15. A pair of wolves on shore ice, Isle Royale. Wolves such 
as this alpha pair regularly travel on the ice to avoid deep snow and 
to seek prey on islets. Photograph: R.O. Peterson, Isle Royale 
National Park archives. 



1992). Blood assays, however, have indicated that 
both disease and inbreeding may be factors in the 
population decline (Wayne et al. 1991). Some 
wolves sampled on Isle Royale had positive, albeit 
low, titers for Lyme disease, and this disease 
appears to cause reproductive problems in some 
mammal species (ibid). Genetic losses of the 
magnitude observed in Isle Royale wolves could 
explain their current low reproductive success 
through inbreeding depression, although the precise 
mechanism is unknown (Wayne et al. 1991). 

It is difficult to predict whether and how soon Isle 
Royale wolves would switch to, or actively seek 
out, introduced caribou, or whether they would 
continue to rely on moose (A.T Bergerud, pers. 
comm.). Wolves search optimally for prey — 
spending search time where prey are most likely to 
be encountered (Bergerud 1985, Bergerud et al. 
1990, Bergerud and Page 1987, Bergerud and Elliot 
1986). When caribou are rare, highly dispersed, or 
inaccessible across water, wolves continue to seek 
more numerous moose even though they are indi- 
vidually more difficult to kill (ibid). On Isle Royale, 
wolf response would depend on individual wolf or 
pack habits, pack travel patterns, and the demo- 
graphics and condition of the moose. Despite 
extraordinary wolf densities, moose densities have 



Habitat Suitability 



21 



persisted at 1-3/km 2 — some three to five times 
higher than most mainland areas in North America 
(Peterson 1990). 

Despite these uncertainties, Bergerud (pers. comm.) 
predicts that wolves will restrict caribou to secure 
escape habitat on Isle Royale. The net effect of high 
wolf numbers (10/1,000 km 2 ) is to greatly reduce 
the carrying capacity for caribou on Isle Royale. 

Where caribou range overlaps with moose and wolf 
ranges, as in northern Ontario and Manitoba, 
caribou densities average at most 0.0 1-0.1 /km 2 , 
equivalent to 5-54 animals on Isle Royale, com- 
pared to 0.3-0.4 caribou/km 2 where moose and 
wolves are less numerous (see review in Appendix 
D). Based on the densities of caribou on occupied 
islets in Lake Nipigon adjacent to high wolf densi- 
ties, 1.2-1.8/km 2 , Isle Royale's islets might support 
about 10-20 caribou (Appendix D). Bergerud (pers. 
comm.) predicts that only 20-55 caribou could 
survive on Isle Royale with wolves present. 

Another potential caribou predator on Isle Royale is 
the lynx. Historically, lynx were common on Isle 
Royale but disappeared in the 1930s due largely to 
overtrapping (Martin 1988). Evidence of natural 
immigration included credible lynx sightings in 
1963, 1970, 1981, and 1988; each being a year 
when regional populations were on the rise and 
irruptive migrations were recorded on the mainland 
(Martin 1988; R.J. Krumenaker, pers. comm.). 
These sightings of individual lynx do not necessarily 
mean that lynx will eventually recolonize and 
reproduce to a viable population size. If lynx do not 
become reestablished on their own, the National 
Park Service will consider restoration actions (R.J. 
Krumenaker, pers. comm.). 

Important lessons can be learned from the Maine 
caribou reintroduction program, where the first two 
years of releases suffered high predation mortality 
in addition to disease problems stemming from the 
captive herd (B. Connery, pers. comm.). The Maine 
investigators concluded that these predation risks 
could have been overcome by "flooding" the release 
area with animals until an initial herd became 



established with the knowledge required to avoid 
predators (ibid). Since the Maine project was 
terminated, the final results of that experiment will 
not be forthcoming. Any plans to release caribou on 
Isle Royale would be experimental as well, since we 
cannot fully anticipate the potential for successful 
restoration on an island archipelago with high wolf 
densities. 



Winter Ticks 

During the last few years Isle Royale moose have 
suffered significant mortality from heavy winter tick 
(Dermacenter albipictus) infestations (Peterson 
1990). Across North America, moose die-off s have 
been linked to winter ticks (Glines and Samuels 
1989). Winter ticks were historically a white-tailed 
deer parasite, to which other cervids are less 
resistant (W. Samuels, pers. comm.). They spread 
northward with white-tailed deer after the 1 860s, 
and their range is still extending northward (Samu- 
els 1989). High rates of winter tick infestations are 
associated with high moose densities and warm, dry 
spring and fall weather (Welch et al. 1990). 

Winter ticks have been collected from a few wild 
woodland caribou in Alberta. Two captive reindeer 
harboring extremely high tick densities died (Welch 
et al. 1990). Although tick infestations have not 
been observed in woodland caribou populations, 
Welch et al. (1990) predicted that winter ticks could 
become a problem for woodland caribou if warm, 
dry weather conditions continued as in recent years. 
On Isle Royale, caribou would be exposed to winter 
ticks due to high moose densities and heavy winter 
tick infestations on moose (Peterson 1990). Poten- 
tially, tick-related caribou mortality could be 
substantial and additive to predation mortality (W. 
Samuels, pers. comm.). 



22 



Results and Discussion 



Escape Habitat 

Categories of potential escape habitat for caribou 
include islands, shorelines, rocky cliffs, and open 
bog wetlands. All of these sites afford the caribou 
greater vision of approaching predators than closed 
forest habitats. Also, caribou readily flee into the 
water or sometimes up escarpments or over ice to 
escape from intruders (see Appendix D). Islands 
and open wetlands separate caribou from predators 
that are reluctant to swim in pursuit of uncertain 
prey resources. 

Of these categories, offshore islets are the most 
likely to provide secure havens for caribou on Isle 
Royale, except during winter when the water 
between the main island and some or all islets 
freezes. The main island of Isle Royale might 
provide safe sites for a smaller number of caribou 
and might supply critical habitat in winters with 
extensive ice on Lake Superior. Due to a lack of 
concentrated winter food supplies, Bergerud (pers. 
coram.) predicts that caribou would select winter 
habitat on Isle Royale based on visibility and snow 
depth to assure their escape flight potential. A 
description of each escape habitat type follows. 

Low Wolf-Use Habitats. In Pukaskwa National 
Park, caribou generally avoid high wolf-use areas. 
Woodland caribou also have been observed to 
"hide" between pack territories or move away from 
a territory where wolves have detected them and to 
"hide" until detected again in a different territory 
(Bergerud 1984, 1989). On Isle Royale, pack 
territories have shifted enough between years (R.O. 
Peterson, pers. comm.) so that caribou probably 
would not be able to predict relatively safe zones 
between wolf packs. 

The only places on Isle Royale that wolves have not 
frequented consistently in recent decades are 
Siskiwit swamp and the 1936 burn area, where 
moose are uncommon (R.O. Peterson, pers. comm.) 
(Figure 9). However, these areas may not provide 
caribou with adequate long-term isolation from 
wolves, given the wolves' ability to roam and adapt 
to new prey resources. 



Wetlands. In northern boreal forests, woodland 
caribou find refuge in large bogs or wetlands with 
small water openings (Bergerud 1978, Abraham et 
al. 1990). On Isle Royale, wetlands are scattered 
across the main island and surrounding islets; 
however, they are typically small or linear and often 
wooded with black spruce or white cedar. Wetland 
forest types total only about 1 ,400 ha (Hansen et al. 
1973). When wolf densities are high, Isle Royale's 
wetlands would probably not be large enough to 
protect caribou from predation except possibly 
Siskiwit swamp, which wolves rarely enter in winter 
(R.O. Peterson, pers. comm.). 

Rugged Terrain. On Pic Island, in Lake Superior, 
caribou escape from predators and human intruders 
by running up steep cliffs and by fleeing into water 
(Ferguson et al. 1988). Rocky cliffs are characteris- 
tic of northwest-facing slopes on Isle Royale. Along 
the north shore of Isle Royale, including many of its 
bays and inlets and most offshore islets, these cliffs 
are at or near the shoreline. Hence, introduced 
caribou might learn to use cliffs as an alternative 
escape route. 

Shorelines. The shape of Isle Royale and its islets — 
long and narrow, striated with long, narrow bays, 
inlets, and lakes — produces a high shoreline:area 
ratio favorable to woodland caribou (Bergerud et al. 
1990) (Figure 2). Caribou released onto Isle Royale 
would most likely wander the shoreline until they 
settled into favorable sites (A.T Bergerud, pers. 
comm.). Secure calving sites might be found along 
stretches of the main island shore where a narrow 
shelf of land is backed by steep cliffs. 

In all but the coldest, and especially the calmest 
winters, parts of the Isle Royale shoreline remain 
open except for narrow shelf ice (R.O. Peterson, 
pers. comm.). When many offshore islets are frozen 
in, prevailing winds will often keep ice from form- 
ing along exposed shorelines of the main island. 
Only once a decade or so does the wind remain calm 
enough for ice to completely encircle Isle Royale for 
a few days or longer (R.O. Peterson, pers. comm.). 
Thus, shorelines on the main island might provide 
winter habitat for caribou. 



Habitat Suitability 



23 



Table 2. Offshore islets, Isle Royale National Park (vegetated islets >0.5 ha). 







Size (hectares) 


Islet Group 


Number of Islets 


Mean 


Range 


Rock and Tobin harbors 


37 


6.2 


0.5-61.0 


North Shore 


23 


11.6 


0.5-181.2 


Malone Bay 


9 


9.1 


1.2-46.9 


Washington Harbor 


7 


34.3 


2.1-127.2 


Houghton Point 


6 


4.8 


1.5-18.9 


Todd Harbor 


5 


1.4 


0.5-2.7 


All Islets 


87 


9.9 


0.5-181.2 



Note: Vegetation cover determined from aerial photographs; area digitized from topographical maps at the Michigan Techno- 
logical University GIS Laboratory. 



North Shore 



Rock Harbor 



Todd Harbor 




Tobin Harbor 



N 

1 



Washington 
Harbor 



Houghton Point 



Moose Density 

lf\ low 

bd moderate 

EH high 



Figure 16. Major offshore islet groups, Isle Royale. Relative moose densities are indicated by islet group, as 
described by R.O. Peterson (pers. comm.). 



24 



Results and Discussion 



Islets. When wolves are present, 
island archipelagoes are considered 
the best sites for reintroducing caribou 
into the southern portions of their 
former range (Bergerud 1980). "Here, 
the releases of caribou could 'island 
hop' to escape wolves." (Bergerud 
1980:99). More than 200 islets 
surround Isle Royale, including 87 
vegetated islets 0.5 ha or greater in 
size (Table 2). Two-thirds of these 
islets are on the eastern end of Isle 
Royale in Rock Harbor and Tobin 
Harbor and along the north shore 
(Figure 16). Other groupings or 
chains of islets are at Todd Harbor on 
the northwest shore, Washington 
Harbor on the west end, and Malone 
Bay and Houghton Point on the south 
shore. Four of the large inland lakes, 
Siskiwit, Desor, Richie, and Sargent, 
contain a total of eight large islets. 




Figure 17. Rock Harbor channel in winter with extensive ice cover. 
When winds are calm during the winter, ice forms between the main 
island of Isle Royale and its surrounding islets. Islets farther from the 
main island shoreline rarely freeze in, perhaps only once in 10 years, 
while closer islets such as Caribou and Mott islands (foreground) are 
connected to the shore most years. Photograph: R.O. Peterson, Isle 
Royale National Park archives. 



The 87 islets average 9.9 ha in size (0.5-181.2 ha), 
with a combined area of about 8.9 km 2 , or less than 
2% of the park land area. Five islets, Amygdaloid, 
Mott, Thompson, Washington, and Wright islands, 
are 0.5 km 2 or larger, and another 15 islets range 
between 0. 1 and 0.5 km 2 , half of which are in Rock 
Harbor and Tobin Harbor. These figures exclude 
Passage Island, a 0.82-km 2 island located 5.5 km 
off the northeast tip of the main island, because it is 
far enough out to discourage caribou crossings from 
Isle Royale. Moose have not colonized Passage 
Island, and a draft NPS policy calls for excluding 
ungulates to protect rare plant resources on Passage 
Island (R.J. Krumenaker, pers. comm.). 

The islets surrounding Isle Royale are typically 
arrayed in long chains paralleling the main island. 
Hence, except for the chain of islets off Houghton 
Point that defines Siskiwit Bay, most of the islets 
are within 1 km of the main island shore. For 
example, the 37 vegetated islets in Rock Harbor and 
Tobin Harbor are, on average, only 430 m from the 
main island shore (range = 24-1,561 m). Since these 
islets are close to each other, the distance to the 



nearest land is much shorter (mean = 49 m, n = 
37, range = 12-171 m) than is the distance to the 
main island. 

Distance to shoreline and exposure to prevailing 
winds are the keys to how frequently and persis- 
tently the different islets freeze in (Figure 1 7). The 
most sheltered islets are locked in with ice by late 
December in most years, while islets that are more 
distant from the main island shore are infrequently 
icebound. Wind patterns have the greatest influence 
on where and when ice forms around Isle Royale. 

Moose as an Indicator of Islet Habitat Quality. 
Bergerud (pers. comm.) recommended assessing 
moose productivity on Isle Royale's offshore islets 
to determine their relative suitability for caribou. 
The general suitability of islets as winter refuge 
habitat is described in Stephens and Peterson's 
(1984) comparison of winter moose densities, calf 
percentages, and wolf kill rates between a group of 
27 of Isle Royale's larger islets (>10 ha) and the 
main island. Winter moose densities were four times 
higher on these islets than on the main island; also, 



Habitat Suitability 



25 



the highest-use areas on the main island were within 
2 km of the shoreline. In January and February, calf 
percentages were nearly twice as high on islets as 
the main island. 

This wintertime concentration of calves on islets, 
however, was a residual from the high proportion of 
calves actually born on islets. In winter, calves 
were killed at disproportionately high rates on islets, 
presumably because wolves knew that they were 
more likely to find calves on the larger islets than 
they were to find them dispersed on the main island. 
Stephens and Peterson (1984:23) concluded that the 
moose's preference for islets in midwinter was 
"largely attributable to the survival advantage that 
accrues to very young calves" in their first summer. 
In another study, restricted to the northeast end of 
Isle Royale, Edwards (1983) demonstrated that 
moose cows with calves concentrated on small islets 
(all islets studied were <10 ha) and along the 
shoreline of Isle Royale and remained near the lake 
shore from May through August. 

The winter density of moose varies greatly between 
offshore islets, correlating loosely with the fre- 
quency of ice formation, islet size, and forage 
resources (Snyder and Janke 1974; R.O. Peterson, 
pers. comm.). For example, Snyder and Janke 
(1974) found that in the Rock Harbor islet chain, 
the forest on Mott Island was significantly less 
browsed than similar forest types on the main 
island, and the forest on Smithwick Island — farther 
out on the chain — was barely browsed at all, 
presumably due to inadequate ice for winter cross- 
ings (Figure 15). 

Peterson's (pers. comm.) summary of relative 
winter moose densities, generalized from his 
observations during the last two decades, provides 
an index to the quality of winter habitat for caribou 
by islet group (Figure 16). Wolves regularly cross 
the ice to access islets with heavy or moderate 
moose densities — especially the larger islets in 
Washington Harbor, Malone Bay, the west end of 
Rock Harbor, and Amygdaloid Island. At the 
opposite end of the spectrum, moose are largely 
absent from, and wolves rarely or never, visit the 



Houghton Point islets and the east end of Rock 
Harbor and Tobin Harbor. These latter islets would 
be most suitable for caribou. 

Summary of Islet Habitat. In summary, the small 
islets surrounding Isle Royale would provide 
summer escape habitat for caribou. Caribou re- 
maining on islets would have a safe haven from 
wolves for at least 9-10 months of the year. Current 
forage supplies, especially herbaceous plants, 
deciduous shrubs, and arboreal lichens, appear to be 
adequate to support caribou at typical island 
densities (A.T Bergerud, pers. comm.). If the islets 
became heavily browsed, presumably, caribou 
would be able to survive on depleted food supplies 
much as they do in Lake Nipigon, the Slate Islands, 
and Pic Island (as long as they could escape preda- 
tion during midwinter). If all of the islets and 
exposed shorelines froze in solidly, however, then 
caribou would not have sufficient escape habitat 
and wolves could kill the entire herd. 



Summary 

Based on this precursory analysis, forage resources 
and spring-to-early- winter escape habitat appear to 
be adequate for a small herd of caribou on Isle 
Royale. The bottleneck for caribou carrying capac- 
ity, and mere survival, would be midwinter escape 
habitat. For caribou to survive on Isle Royale with 
wolves present, a low frequency of complete ice 
formation and the availability of alternative escape 
habitat when ice was widespread would be required. 
Isle Royale would clearly be most suitable for 
caribou if and when the wolf population was at low 
numbers, but wolf recovery would eliminate caribou 
from most land areas of the park. In the short-term, 
survival of reintroduced caribou would depend on 
how quickly they learned to use available escape 
habitat and the luck of mild winters while the herd 
became established. The probability of longer-term 
persistence is evaluated in the following section. 



26 



Results and Discussion 



Population Vulnerability 
Analysis 

Population vulnerability analysis is the process of 
estimating minimum viable population (MVP) sizes 
for specific populations (Gilpin and Soule 1986). A 
minimum viable population is the threshold number 
of animals required to ensure, at some defined level 
of risk, that a population will persist for a given 
time interval at a particular location. Conventional 
standards for minimum viable populations 
include (1) greater than 90% certainty of long-term 
persistence, usually centuries, (2) population 
maintenance in nature without significant demo- 
graphic or genetic manipulation, and (3) retaining a 
replacement level of immediate fitness — vigor, 
fertility, and fecundity — with sufficient genetic 
variation to adapt by natural selection to changing 
environments (Soule 1987). 

In current theory, four kinds of variation indepen- 
dently threaten small populations. These variations 
are (1) demographics or birth-and-death processes, 
(2) genetics, (3) environmental stochasticity, and (4) 
catastrophes (Soule 1986). Population fragmenta- 
tion or metapopulation structure, especially systems 
based on local extinctions and recolonizations, also 
contribute to MVP estimation (Gilpin 1987). When 
these elements interact, an extinction vortex, or 
downward spiral in population size, can result 
(Gilpin and Soule 1986). 

No single, universal minimum viable population can 
guarantee that animal populations will persist 
(Gilpin and Soule 1986). However, based on theory 
and empirical observations, ballpark numbers or 
orders of magnitude of individuals required to 
maintain minimum viable populations of completely 
isolated populations have been estimated. For 
species with "average" reproductive rates and 
genetic variability, including wild caribou, a popu- 
lation of at least several hundred individuals may be 
necessary to establish a minimum viable population 
CLande and Barrowclough 1987). 



Thus, we can predict that a fluctuating population 
of around 200 caribou in the absence of moose and 
wolves, such as existed historically on Isle Royale, 
could not persist indefinitely in isolation from other 
populations. A population of only 20-55 animals, as 
Bergerud predicts for reestablishment on Isle 
Royale, would have a low probability of surviving 
for many decades, much less centuries. 

More accurate and informative predictions can be 
made by simulating the growth of an introduced 
population with a computerized model, which I did 
for a hypothetical Isle Royale caribou herd using 
Robert Lacy's (1990) VORTEX model (Appendix 
B). The VORTEX model uses Monte Carlo simula- 
tions of demographic events, environmental varia- 
tion, and catastrophes to calculate population 
persistence times. Such modeling permits predic- 
tions of how long populations established by 
different reestablishment schemes would survive, or 
whether the effort is worthwhile. For these simula- 
tions, the number of animals released, number of 
years of releases, mortality rates, and carrying 
capacities were varied between simulation runs. 



VORTEX Results 

In preliminary trials with high carrying capacities, 
so that population size was not immediately trun- 
cated below the number of animals initially re- 
leased, we determined that persistence times were 
not improved by releasing more than about 75 
animals or by extending the release time over more 
than one year. Subsequent simulations with lower, 
more realistic carrying capacities reduced to three 
basic variants: carrying capacity equals 54 with 
high and moderate mortality, and carrying capacity 
equals 27 with high mortality (Appendix E). 
Unfortunately, I was not able to satisfactorily model 
two release options that might overcome the initial 
effects of high wolf predation: "swamping" the 
release site with woodland caribou, and releasing 
caribou while wolves were absent and subsequently 
restoring the wolf population (i.e., gradually 
increasing wolf predation rates as wolf densities 



VORTEX Results 



27 



increased). To summarize, a translocated caribou 
population with fewer than 54 animals and suffering 
high mortality would not survive for 50 years 
(Table 3). The mean time to extinction was about 
33-44 years. With less severe mortality and a larger 
mean population size of 54 animals, mean persis- 
tence time would increase to 79 years. 



Lessons from Elsewhere 

Translocating small numbers of ungulates into 
favorable habitats has frequently succeeded in 
establishing large populations. Griffith et al. (1989) 
found that for native game species, just 20-40 
founding animals were sufficient to predict high 
translocation success. Success was defined only as 
attaining a "self-sustaining population" — long-term 
persistence was not addressed. Some caribou herds 
have prospered from initial transplants of fewer 
than 20 animals into favorable habitats (see history 
of caribou transplants in Appendix A). 

A previous attempt at modeling caribou reestablish- 
ment alternatives was part of the Maine caribou 
reintroduction plan (McCollough 1987). A stochas- 
tic model based on Grier (1980) was employed to 
evaluate six release options for three projected 
levels of herd survival and fecundity. The Maine 
simulation predicted only short-term probabilities of 



achieving the approximately 100 animals believed 
to be necessary for long-term viability (M. McCol- 
lough, pers. comm.). These simulations revealed 
that only restorations of herds with initial high 
survivorship and intermediate or high fecundity 
exceeded the target herd size of 100 animals in 10 
years. Releases extending over 5 years and totaling 
98-1 15 animals with heavily skewed adult sex ratios 
performed better than 3-year releases totaling less 
than 75 animals. The greatest determinant of 
success was population survival rate, especially for 
yearling-only releases. 

As it turned out, mortality in the two years of actual 
releases of woodland caribou to the Baxter State 
Park region of Maine far exceeded the modeled 
scenarios, due principally to black bear predation 
and diseases carried from captivity (McCollough 
and Connery 1991). Based on this experience, the 
advisory committee of the Maine project concluded 
that three releases of at least 40-60 animals would 
have been necessary to overcome initial high 
mortality rates and establish a core herd of animals 
acclimatized to the new habitat. 

Several small populations of woodland caribou 
persist along the north shore of Lake Superior, 
which could lead to the misleading conclusion that 
small populations may be viable. These herds have 
been isolated for only 15-30 years and stray bulls 
have been recorded moving between some north 



Table 3. Simulated population persistence times (in years) of selected release scenarios for woodland 
caribou on Isle Roy ale (50 simulations per 100-year run). 



Runs 1 


Mortality 
Rate 


Carrying 
Capacity 


Persistence Time (in Years) by Percent 
of Populations Surviving 


90% 50% 10% 


1 
2-3 
4-6 


High 

Moderate 

High 


27 
54 
54 


21 31 39 50 
63-64 79-80 93-94 >100 
31-39 44-50 55-61 70-72 



1 Runs are combined by mortality rate and carrying capacity; number of animals released and occurrence of catastrophes had 
minimal influence on the results. Number of animals released: run 2 = 52, runs 3-5 = 78, run 6 = 104. Catastrophes were 
omitted from run 5. See Appendix E. 



28 



Results and Discussion 



shore herds in the last 15 years (Bergerud 1985, 
1989a; Darby et al. 1989). Furthermore, the 
prognosis for all the Lake Superior herds is bleak. 
Bergerud (1989a), for example, predicted a high 
probability of extinction for the Pukaskwa herd 
within 25 years. Fluctuations in the size of the Slate 
Islands herd, characterized by periodic widespread 
starvation, have been increasing in amplitude (A.T. 
Bergerud, pers. comm.). This fluctuation suggests 
that total population collapse is possible despite a 
mean population size of 250-400 animals. Thus, I 
conclude that historic Lake Superior populations, 
including Isle Royale, were probably part of a 
regional metapopulation with regular gene flow 
between groups and ready recolonization of islands 
from the mainland. 



Summary 

The prospects for caribou restoration on Isle Royale 
can be viewed on two time scales. In the short run, a 
small release of 1 2-24 animals might be sufficient to 
establish a nuclear herd given adequate escape 
habitat and assuming wolves are present. If the 
released caribou learned how to avoid wolves (find 
and exploit secure habitat), then initial population 
growth should be good. But as the herd grew it 
would be subject to high mortality rates — 10-22% 
annual adult mortality — and the population prob- 
ably could not exceed a density of 0. 1/km 2 (K = 
54), with wolves numbering six or more. Under 
these conditions, the VORTEX model predicts that 
the caribou herd would not survive for 40-60 years. 

When a proposed restoration site is too small to 
support a minimum viable population at carrying 
capacity, restoration may be possible by managing 
several small populations as a single "metapop- 
ulation." A metapopulation consists of a constella- 
tion of small subpopulations that interact loosely 
through differential dispersal, establishment, 
growth, and extinction (Gogan and Cochrane 1994). 
Managing several small populations as a metapopu- 
lation could involve providing movement corridors 
for individuals to naturally disperse among subpop- 



ulations and to directly relocate, as well as includ- 
ing captive individuals at zoological parks as a 
subpopulation (Gogan 1990). 

I conclude that releasing a reasonable number of 
caribou, approximately 100 animals, would not 
result in an independently viable population on Isle 
Royale or any population at all without follow-up 
releases beginning as soon as 10 years. This result 
is not surprising based on MVP theory and given 
assumptions of high mortality rates, high variance 
in mortality, and a carrying capacity of no more 
than 54 animals. Historically, caribou on Isle 
Royale did not have to contend with wolves, and 
movements to and from the mainland would have 
prevented inbreeding and recolonized the island 
after infrequent catastrophic declines. Replicating 
the natural functions of this historic link to main- 
land caribou populations would be essential to any 
restored Isle Royale population. Hence, as long as 
wolves are present on Isle Royale, successfully 
restoring woodland caribou may only be possible by 
managing a regional metapopulation. 



Reestablishment Methods 

The goal of this section is to evaluate whether 
translocation of caribou to Isle Royale is feasible. 
Analysis of the logistics, costs, and relative benefits 
of specific alternatives is left to the restoration plan 
and environmental assessment if the NPS pursues 
the project in the future. 

The Maine woodland caribou reintroduction plan 
(McCollough 1987), Selkirk Mountains, Idaho, 
herd supplementation plan (Summerfield 1985a, 
1985b), and subsequent program reports (McCol- 
lough and Connery 1990; Servheen 1988, 1989) 
provide a thorough review of alternative caribou 
capture and release procedures. Release options can 
also be gleaned from a proposal for reintroducing 
caribou to Minnesota (Karns and Lindquist 1986). 
Recent translocation projects in Ontario add to the 



Reestablishment Methods 



29 



wealth of examples from which Isle Royale could 
draw, especially in transporting caribou between 
Lake Superior islands. 



Alternative Reestablishment 
Strategies 

Reestablishment strategies vary principally in how 
long the animals are held before release and whether 
they are caught in the wild or captively reared 
(Summerfield 19856; McCollough 1987; McCol- 
lough and Connery 1990; G. Eason, pers. comm.). 
Caribou may be raised in nursery herds either with 
adult animals or by humans. Direct translocation 
approaches include ( 1 ) capture and release of 
calves only, (2) delayed release with holding for a 
few months, (3) quick release with holding for a few 
days to three weeks, and (4) immediate release. 
Target sex and age ratios for releases depend upon 
initial project goals. Population growth is maxi- 
mized by releasing prime breeding-age animals. The 
average sex ratio in nonhunted caribou herds is 
about two males:three females (Bergerud 1980). 
Caribou are polygynous, however, and a typical 
breeding male: impregnated female ratio is about 1 :5 
(R. Page, pers. comm.). Given the risks of high 
initial male mortality and logistical constraints on 
selective captures, a practical release goal for 
caribou is around one male: two-three females 
(Summerfield 19856; McCollough 1987; G. Eason, 
pers. comm.). 



Release Timing 

In their study of translocation success rates, Griffith 
et al. (1989) found that multiyear releases were not 
significantly more successful than onetime releases. 
In our simulations, multiyear and single-year 
releases of the same number of animals performed 
similarly. 



The best time to release caribou and presumably to 
capture them for quick or immediate release strate- 
gies depends on whether deterring dispersal is the 
major goal, or if the caribou would gain by learning 
the new terrain at particular seasons. Logistical 
constraints might dictate release timing, such as 
when the animals could be caught at the source area 
and when transportation is available. Some release 
times used by other projects include (1) early fall/ 
prerut when animals are in good condition and 
congregated in groups, (2) early winter/postrut 
when females are pregnant and dispersal from the 
release site may be limited by snow, (3) late winter 
before females are ready to calve and when caribou 
may be using more open habitat to facilitate cap- 
ture, and (4) spring, just before calving, to allow 
calving in and summer adjustment to the new 
location. 

A few techniques have been suggested to help retain 
caribou at release sites and to deter wolf predation 
during the first year (A.T Bergerud, T. Kreeger, 
R. Page, L. Rogers, pers. comm.). These techniques 
include (1) salt licks, (2) penning, (3) chasing 
swimming caribou back to islets, (4) stationing 
monitoring staff at the release site over winter, (5) 
hand-rearing calves at the release site, and (6) 
conditioning the caribou with dogs to be wary of 
predators. 



Disease and Parasite Screening 

Since wild caribou would be translocated from 
Canada, U.S. Department of Agriculture (USDA) 
requirements for live animal importation and U.S. 
Fish and Wildlife Service permit regulations would 
have to be met (Summerfield 19856; McCollough 
1987; S. Schmidt, pers. comm.). These regulations 
require testing for tuberculosis (TB) and brucello- 
sis, and possibly Bang's disease, before animals 
may be shipped across the international border. 
Specific USDA requirements are determined in each 
case by the state veterinarian and the USDA 
Washington office. Since the TB test takes 72 
hours, animals have to be held at the capture site, a 



30 



Results and Discussion 



source of stress and potential mortality. Wild 
caribou and moose are not known to carry TB, and, 
in fact, the standard TB testing procedure was 
developed for cattle and may not be valid for 
cervids (S. Schmidt, pers. coram.). 

Caribou captured in Newfoundland for rearing and 
release in Maine were held for exhaustive disease 
and parasite testing (McCollough 1987, McCol- 
lough and Connery 1990). The Maine project's 
primary concern was to prevent potentially import- 
ing Elaphostrongylus cervi, a nematode parasite of 
European reindeer. This parasite infects caribou 
herds in central Newfoundland. 

In the Idaho translocation program, captured 
animals were treated with a single dose of Ivermec- 
tin as a general worming agent at the same time 
they were tested for TB and brucellosis (G. Serv- 
heen, pers. comm.). E. cervi has not been detected 
in the Canadian source herd, hence, no treatment or 
quarantine was prescribed for that parasite. 



Sources of Caribou — the Slate 
Islands 

The closest source of caribou for transfer to Isle 
Royale is the Slate Islands herd in Ontario, 112 km 
northeast of Isle Royale. Because of the high 
density of caribou on the Slate Islands, this herd has 
been used for transplants around Lake Superior in 
Ontario (Appendix A). Ontario Ministry of Natural 
Resources policy (WM.2.10.U5, dated 10 August 
1988) permits transferring live animals out of 
Ontario for legitimate and well-documented reestab- 
lishment programs. Removing caribou from the 
Slate Islands to restore extirpated populations 
elsewhere is approved in the Slate Islands Provin- 
cial Park management plan. Beyond the Slate 
Islands, the nearest source of caribou for transfer at 
this time is either Manitoba or far northwestern 
Ontario (H.R. Timmermann, pers. comm.), which 
would be an exponential increase in transport time 
and costs to Isle Royale. 



Caribou have been captured on the Slate Islands 
with permanent drive traps and bait traps (G. 
Eason, pers. comm.). Before transport, the caribou 
have been sedated with Atravate, then tied to 
stretchers and flown by Twin Otter float plane 
directly to the release sites (Figures 18 and 19). A 
maximum of six animals were transported at a time 
with a total flight time of less than two hours. The 
entire capture and transport effort required four to 
eight people. Since disease testing was unnecessary 
for transfers within Ontario, the animals were 
released immediately and generally fared well (G. 
Eason, H.R. Timmermann, pers. comm.). Recent 
transfers off the Slate Islands were done in the fall, 
primarily because government planes were available 
then (G. Eason, pers. comm.). Long-distance 
dispersal was not an issue for the Canadian island 
release sites and was not a major problem for the 
1989 release on the Gargantua Peninsula (G. Eason, 
pers. comm.). 

Potential drawbacks of transferring animals from 
the Slate Islands herd would include inbreeding, 
nutritional stress, and no predator wariness. Due to 
isolation, the animals are inbred, with reduced 
heterozygosity and fewer alleles than North Ameri- 
can averages for caribou (A.T Bergerud, pers. 
comm.). Delayed breeding and delayed antler 
growth in Slate Islands caribou may be genetically 
linked (A.T. Bergerud, pers. comm.), but these 
conditions may also result from prolonged nutri- 
tional stress. Females from the Slate Islands re- 
leased into better environments have reproduced 
normally (G. Eason, pers. comm.). 

Another concern about Slate Island's caribou is that 
they lack conditioning for escaping predators. This 
lack of conditioning may have contributed to high 
predation losses in the first months after release on 
the Gargantua Peninsula, as indicated by caribou 
that left the offshore islands (where they had been 
released) for vulnerable mainland sites (G. Eason, 
pers. comm.). However, Slate Islands caribou are 
still wary of humans and flee readily to water (H.R. 
Timmermann, pers. comm.). Thus, Slate Islands 



Reestablishment Methods 



31 




Figure 18. Capturing woodland caribou on the Slate Islands. 
Caribou are captured on the Slate Islands singly in traps baited with 
salt. They are then blindfolded, sedated, and tagged or radio-collared 
before transport. Photograph: H.R. Timmermann, Ontario Ministry of 
Natural Resources. 




Figure 19. Preparing caribou for transport off the Slate Islands. In 
the first Slate Islands translocations, shown here, caribou were loaded 
into crates and transported by barge. Recently, transportation has 
been improved by placing the animals on stretchers and moving them 
by float plane. Biologists ride with the animals — no more than six at a 
time — and massage their legs during the one-hour flight to maintain 
blood circulation and keep the animals calm. Photograph: H.R. 
Timmermann, Ontario Ministry of Natural Resources. 



caribou apparently have the basic 
predator avoidance instincts but need 
to learn site-specific escape habitats. 

Freutel and Lankester (1989) found 
nine species of gastrointestinal 
parasites during extensive tests on 
Slate Islands caribou, most of which 
are host-specific to caribou (M. 
Lankester, pers. comm.). Of the four 
Elaphostrongyline species, only 
Parastrongylus andersoni has been 
found in Slate Islands caribou, and 
then only in 4% of the herd (Lankester 
and Hauta 1989). P. tenuis and E. 
cervi are most certainly absent (ibid). 

Thus, the Slate Islands caribou appear 
to be comparatively free of contagious 
diseases and parasites that could 
spread to other animals. Yet additional 
testing would be required to assess 
any potential risks to Isle Royale 
moose posed by introduced Slate 
Islands caribou, particularly from P. 
andersoni. Caribou captured for 
transport to Isle Royale would have to 
be screened and possibly treated for 
parasites and diseases before release. 

One final caveat: The Slate Islands 
herd is not immune from extirpation, 
even within the next decade or two. 
Bergerud (pers. comm.) predicts that 
the caribou are likely to crash below a 
sustainable level in one of the next 
population cycles. Ultimately, caribou 
cannot persist in the Lake Superior 
region if they are isolated in small 
populations. The Slate Islands herd 
will not persist indefinitely unless it is 
managed as part of a regional meta- 
population. 



32 



Results and Discussion 



Summary 



Vegetation 



In conclusion, Slate Islands caribou are a relatively 
inexpensive and easily captured nearby source of 
caribou for an Isle Royale transplant. These animals 
would be the logical source for initial releases on 
Isle Royale. Potential problems with the condition, 
wariness, and availability of Slate Islands animals 
would need to be addressed further as part of a 
detailed release plan. Further study of diseases and 
parasites in Slate Islands caribou and their potential 
impacts on Isle Royale moose would be required for 
the plan's environmental assessment. 



Impacts and Management 
Implications 

Potential impacts of caribou on Isle Royale re- 
sources are suggested here to provide guidance for 
an environmental assessment if restoration is 
pursued. I have not attempted to quantify impacts 
such as browse depletion. 

Caribou restoration would prompt changes in a 
number of Isle Royale National Park management 
practices — both temporary and permanent. In 
addition to evaluating and planning caribou restora- 
tion to minimize impacts on other park resources, 
managers would have to consider altering existing 
management programs such as visitor and fire 
management policies to facilitate caribou restora- 
tion. Bergerud ( 1 989a) recommended a number of 
steps to protect caribou in Pukaskwa National Park 
that might be applicable to Isle Royale. 

All but 1 % of Isle Royale National Park's land area 
is legally designated as wilderness. Park manage- 
ment strives to provide a high-quality "wilderness 
experience" for park visitors. Thus, park resources 
potentially affected by caribou include not only 
natural resources — plants, animals, and natural 
systems — but park cultural resources, visitor use 
facilities, and wilderness aesthetics. 



Given a scenario where wolves confined caribou for 
the most part to offshore islets and adjacent shore- 
lines, caribou would have little impact on vegetation 
resources on the main island. In contrast, vegetation 
on occupied islets could be altered significantly, as 
shown by the Slate Islands, Pic Island, and islands 
in Lake Nipigon (Euler et al. 1976, Ferguson et al. 
1988, Bergerud et al. 1990). If 20-50 caribou spent 
much of the year browsing on Isle Royale's islets, 
roughly 9 km 2 (excluding Passage Island) of forage 
supplies would be affected. Through selective 
foraging, the composition as well as abundance of 
islet plant life would be altered. In the absence of 
countervailing influences such as fire, spruce 
budworm outbreaks, or climate change, heavy 
browsing would intensify or accelerate the natural 
successional trend from mixed deciduous forest to a 
more pure conifer forest. 



Endangered Species and Rare 
Plant Communities 

Slavik and Janke (1987) reviewed rare and unusual 
plant occurrences on Isle Royale. Arctic and alpine 
plant communities on the rocky, Lake Superior 
shoreline include numerous postglacial relics, and 
eight state-listed threatened species and two state- 
designated special-concern species. Altogether, 39 
state-listed threatened and endangered species and 
16 special-concern species have been found on Isle 
Royale; for some, Isle Royale is the species' only 
site in Michigan (Slavik and Janke 1987). Caribou 
should not be released onto Isle Royale until their 
likely impacts on these species are considered. A 
rare plant survey focusing on the rock shore com- 
munities of offshore islets would be especially 
relevant to a caribou restoration environmental 
assessment. 



Impacts and Management Implications 



33 



Other Animals 

Competition between moose and caribou for food 
supplies has not been reported, but is possible (A.T. 
Bergerud, pers. comm.). Moose and caribou are 
generally segregated either by food preferences or 
habitat selection (Bergerud and Page 1987, Berger- 
ud and Elliot 1986, Abraham et al. 1990). However, 
on an archipelago like Isle Royale where dispersal is 
not possible and moose are abundant, these two 
cervids could compete for winter foods (Peterson 
1977). 

Introduced caribou would provide wolves with a 
second alternative prey to moose — in addition to 
beaver — that is easier to kill but possibly harder to 
find than moose. Yet, if projected caribou densities 
are correct, total prey biomass would not increase 
much on Isle Royale. Hence, predation rates on 
moose might not change much either. The role of 
caribou as alternate prey for wolves would depend 
on how closely moose and caribou vulnerabilities 
were linked (e.g., if vulnerability of both animals 
was tied to weather, snow depth, etc., or if vulnera- 
bility was based on independent population demo- 
graphics such as moose population age structure). 

Another predator that could be affected by caribou 
restoration is lynx. While presently extremely rare 
on Isle Royale, lynx could increase either naturally 
or through reestablishment actions. Caribou calves 
could provide an alternate prey for lynx during low 
numbers of snowshoe hare (Bergerud 1971). 

Caribou restoration would complicate the decision 
process for wolf reestablishment if the wolves 
become extirpated. Given the current uncertainties 
of wolf recovery, whether natural or artificial, a 
caribou restoration plan would have to conserva- 
tively assume that wolves would be present on Isle 
Royale in the long run. In other words, a realistic, 
worst-case scenario for caribou on Isle Royale 
includes high wolf numbers. 

The current record-low number of wolves and 
projections of additional decline would appear to 
foster caribou restoration on Isle Royale in the short 



run. The probability that caribou reestablishment 
would succeed increases if it was delayed until after 
wolves had declined to five or fewer animals. 
However, given the potential for wolf population 
recovery, the long-term prospects for a self-sustain- 
ing caribou population appear to be poor as long as 
wolves are present. 



Diseases and Parasites 

The potential for introducing new diseases and 
parasites to Isle Royale's isolated animal popula- 
tions must be a major concern for any proposed 
introductions. While caribou were once native to 
Isle Royale, their absence for more than 60 years 
may have left alternative hosts in the park free of 
pathogens carried by caribou. Also, introduced 
animals may carry different parasites or diseases 
from those carried by Isle Royale's native caribou, 
particularly pests that have invaded or spread to the 
Great Lakes region during this century. Moose 
would be vulnerable to cross-species transfer of 
pathogens from caribou, especially since the Isle 
Royale moose herd has apparently been isolated 
from mainland environments for many decades. 

Lankester and Fong (1989) published a thorough 
review of the distribution of the four species of 
Elaphostrongyline parasites carried by caribou, 
which are P. tenuis, E. cervi, P. andersoni, and P. 
odocoilei. The intermediate hosts for all four 
species are gastropods. Any abundant and mobile 
native species of snails and slugs are probably 
suitable hosts (Lankester and Hauta 1 989). 

The meningeal brainworm is common in white- 
tailed deer, its normal definitive host, across eastern 
North America. The parasite causes severe neuro- 
logic disease in other cervid species. Woodland 
caribou infected with this parasite usually die 
(Anderson and Strelive 1968, Anderson 1971). 
Meningeal brainworm is widespread in Michigan 
white-tailed deer (T Cooley, pers. comm.). A 
previous report of meningeal brainworm larvae 
from Isle Royale moose pellets (Karns and Jordan 



34 



Results and Discussion 



1969) has since been discounted (Lankester and 
Hauta 1989). Given the similarity in appearance, 
the larvae collected on Isle Royale may have been P. 
andersoni but could also have resulted from 
collecting equipment contamination (M. Lankester, 
pers. coram.). 

The muscleworm, P. andersoni, is also common in 
white-tailed deer in North American (Anderson and 
Prestwood 1981, Pybus and Samuel 1984) and 
woodland caribou in Labrador and Ontario (Lank- 
ester and Hauta 1989). P. andersoni and another 
muscleworm, P. odocoilei, found in caribou and 
deer in western North America, are not considered 
lethal cervid parasites, although they can cause 
verminous pneumonia and severe myositis in their 
hosts (M. Lankester, pers. coram.). The potential 
impact of P. andersoni on moose is unknown. Since 
moose on Isle Royale are already heavily infested 
with both hydatid tapeworms (Echinococcus 
granulosus) and winter ticks, they may be espe- 
cially vulnerable to other new parasites (R.O. 
Peterson, pers. comm.). 

Introduced to North America through reindeer 
imported to Newfoundland, Elaphostrongylus cervi 
rangiferi also causes neurologic disease in caribou. 
Experimental infestations of moose with£. cervi 
caused pathological changes and paralysis (Lank- 
ester 1976). A parasite resembling E. cervi was 
identified in woodland caribou in Ontario (Lank- 
ester 1976, Lankester and Northcott 1979). Subse- 
quently, this parasite was positively identified as the 
muscleworm, P. andersoni (Lankester and Hauta 
1989). 

The greatest risk to Isle Royale moose is the 
introduction of elophostrongyline nematodes, 
particularly E. cervi and P. andersoni. In North 
America, E. cervi has only been found in introduced 
reindeer and in caribou from Newfoundland. The 
muscleworm P. andersoni is present at low densities 
in Slate Islands caribou, as well as white-tailed deer 
and caribou in some other parts of the Lake Supe- 
rior region (Lankester and Fong 1989; M. Lankest- 



er, pers. comm.). Additional studies should be 
conducted to determine the effects of P. andersoni 
on moose and whether it is present on Isle Royale. 

Slate Islands caribou could also import intestinal 
parasites to Isle Royale (Fruetel and Lankester 
1989). To date, no sampled Slate Islands caribou 
have carried the tapeworm Taenia krabbei, which 
requires the wolf as a final host (M. Lankester, 
pers. comm.). T. krabbei is not present on Isle 
Royale (R.O. Peterson, pers. comm.). The tape- 
worm's introduction should be avoided, even though 
it has little impact on moose populations where it is 
common (M. Lankester, pers. comm.). Finally, 
botfly species carried by caribou are host-specific 
(M. Lankester, pers. comm.), so they would not be 
expected to transfer to moose. 

In summary, extensive parasite and disease screen- 
ing on the Slate Islands or other source herd and 
Isle Royale would be necessary before woodland 
caribou were translocated between these parks. 
Imported caribou would have to be certified free of 
infectious diseases by Canadian and American 
veterinarians before being released on Isle Royale. 
A protocol for parasite testing and, if necessary, 
treatment would have to be included in a caribou 
restoration plan. Risks of parasite and disease 
introduction should be weighed carefully in deci- 
sions about restoration. 



Wilderness 

Intensive browsing would affect the appearance of 
vegetation on offshore islets. However, many 
visitors would not notice until conditions were 
severe; the dramatic effects of moose browse have 
to be pointed out to most summer visitors (pers. 
obs.). Negative reactions might be outweighed by 
the perceived bonus of seeing or knowing about 
caribou in the park. 

Monitoring activities would also impact the wilder- 
ness character of Isle Royale. Radio collars on 
animals might be objectionable to some visitors. 



Impacts and Management Implications 



35 



These concerns were raised when the National Park 
Service agreed to radio-collaring moose in 1983 and 
wolves in 1988, but the debate has remained largely 
philosophical because few visitors actually see the 
collared animals (R.O. Peterson, pers. comm.; pers. 
obs.). 

A much more intrusive element of radiotelemetry 
monitoring is low-level overflights. By regulation, 
float planes cannot land outside of three designated 
sites in Isle Royale National Park, except at the 
discretion of the park superintendent. The National 
Park Service also attempts to restrict flight intru- 
sions over the wilderness by imposing restrictions 
on government and concessionaire flights. 

Caribou monitoring would have to be even more 
intensive than the current wolf telemetry studies to 
provide accurate information on calving and sources 
of mortality. If caribou were concentrated on 
offshore islets, much of the summer monitoring 
might be done from boats. Although more time and 
labor consuming, this option would be much less 
expensive and intrusive than airplane flights. Other 
alternatives to extensive monitoring flights include 
satellite telemetry and attaching antenna to existing 
radio or fire towers. 

The visitor's park experience could also be affected 
by campground and trail closures and other restric- 
tions on boat or foot travel implemented to protect 
caribou. Restrictions on undesignated backcountry 
camping have already been instituted largely to 
protect wolves, without appreciable concern from 
the public (R.J. Krumenaker, pers. comm.). Caribou 
protection would probably be focused on offshore 
islets, most likely on the northeast end and south 
shore, which are already closed to undesignated 
camping. 

Caribou readily become accustomed to routine 
human activities such as logging (A.T Bergerud, R. 
Page, pers. comm.). Moose are not only habituated 
to Isle Royale's campgrounds and settled areas, but 
some cows with calves actually seek out these 
human-use areas as a refuge from wolves and have 
higher calf survival as a result (Stephens and 



Peterson 1984). Thus, other than trying to prevent 
curious visitors from directly harassing caribou, 
"visitor management" for caribou on Isle Royale 
would be primarily educational. 



Summary 

This review highlights potential impacts and 
management concerns with attempting to restore a 
caribou herd on Isle Royale. Any proposal to 
reestablish caribou in the park would require a 
thorough environmental analysis, including quanti- 
tative impact analysis. Studies would be needed to 
illuminate the risks of caribou introducing diseases 
or parasites to Isle Royale's insular moose herd and 
the susceptibility of caribou to winter tick infesta- 
tion. Rare plant distribution and threats from 
caribou trampling and browsing would also have to 
be investigated. Another study topic would be 
visitor attitudes about caribou restoration. 

The scale of a caribou restoration project as pro- 
jected in this study would be unprecedented for 
resource management on Isle Royale. The inter- 
agency, public, and private effort required to plan, 
fund, and implement caribou reestablishment would 
far exceed the ongoing peregrine falcon (Falco 
peregrinus) restoration project and perhaps even the 
moose and wolf research effort. Park managers 
would have to be committed to a lengthy program 
requiring considerable logistical support, regardless 
of the source of project funding and staff. The 
initial translocation would be followed by at least 
three to five years of intensive population monitor- 
ing and require flexibility to deal with complica- 
tions. 

Yet the odds of successful woodland caribou 
restoration on Isle Royale are linked most directly to 
the fate of wolves in the park, and hence, to deci- 
sions about managing the wolf population. In the 
long-term, population viability goals for caribou are 
probably not attainable as long as wolves are in the 
park, hence, "artificial immigration" or periodic 
imports of new caribou would be required to 



36 



Results and Discussion 



maintain a caribou herd. Managing a caribou herd 
under these circumstances would necessarily be 
experimental and does not fit strictly under NPS 
policies for native species restoration (see the 
Introduction). 



Impacts and Management Implications 37 



Conclusions 



We cannot say definitively whether caribou restora- 
tion would succeed on Isle Royale. This report 
shows that conditions on Isle Royale could be 
suitable for caribou — except that with high numbers 
of wolves, escape habitat would not be secure in 
severe winters. We do not know the frequency of 
winters severe enough to eliminate all water-related 
escape havens. More importantly, how caribou 
would adapt to Isle Royale and whether wolves 
would learn to seek out caribou is uncertain. These 
questions could only be answered experimentally. 
But assuming that five or more wolves were present 
(0.01 /km 2 ) and restricted caribou to no more than 
54 animals (0.1/km 2 ), the VORTEX model predicts 
that the herd would persist for less than 40-60 
years. 

Logistically, translocating caribou from Ontario 
would not be difficult. Complications might arise 
from disease and parasite screening and treatment. 
While logistically feasible, the project would be 
expensive. Including the expense of diligent moni- 
toring, the total bill for caribou restoration would 
far exceed previous projects in the park — up to 
$100,000 per year. 

Woodland caribou inhabited Isle Royale for thou- 
sands of years, coevolving with the island's unique 
biotic resources. Yet the island ecosystem has 
changed since caribou were extirpated and caribou 
restored with wolves present are unlikely to behave 
and disperse as they did historically. Thus, potential 
impacts of caribou on present park resources would 
leed further study before restoration could be 
approved. Issues that would need to be resolved 
include the caribou's potential effects on rare plants 
and the effects of caribou diseases and parasites on 
moose. 



Addressing the NPS management policies on page 
1,1 conclude: 

1 . Adequate proof exists that caribou lived on 
Isle Royale and that they disappeared or 
were substantially diminished as a direct or 
indirect result of human-induced change to 
the species and ecosystem. 

2. Adequate foraging habitat exists on Isle 
Royale for caribou. Secure winter escape 
habitat may not be adequate to protect 
caribou from wolf predation during severe 
winters. Also, translocated caribou might not 
learn to use the available escape habitat 
before they were extirpated by wolves. If 
predictions for wolf predation are correct, 
then the caribou would not achieve a self- 
perpetuating, viable population size. 

3. Caribou would not pose a threat to the safety 
of park visitors or to persons or property 
outside park boundaries. They might, 
however, threaten rare plants in the park or 
carry diseases or parasites threatening to 
moose. Further study would be needed to 
assess these potential threats. 

4. Assuming five or more wolves were present 
at the time of release, caribou would likely 
encounter problems with predation either at 
the release site or more likely as they subse- 
quently wandered on the main island. 

5. Caribou attained from Ontario near or in 
Lake Superior would nearly approximate the 
caribou subspecies extirpated from Isle 
Royale. 

6. The prospects for natural (unassisted) 
reestablishment are virtually nil. Restoration 
has a poor chance of success (measured in 
decades) unless either wolves are nearly or 



39 



totally extirpated, or restoration is defined to 
include "artificial," periodic infusions to the 
caribou herd. 

7. A restoration action plan has not been 
developed and is not proposed currently. 

Ironically, the only "natural" system that has been 
proven to work on Isle Royale, by surviving for 
more than 3,500 years of prehistory, is a caribou- 
lynx-snowshoe hare-Native American community. 
Yet it is clear that this "primitive vignette" (Leopold 
et al. 1963) of Isle Royale cannot be replicated 
simply by translocating caribou to the island. The 
chief obstacles to caribou restoration on Isle Royale 
would be wolves and moose — two species that 
arrived on the island in this century as a result of 
human-induced changes on the mainland. 

Barring regular management intervention, the odds 
are high that wolves would eliminate caribou from 
Isle Royale within a few decades. In other words, 
interjecting caribou into the present wolf-moose 
system is unlikely to establish a self-sustaining 
caribou herd. I based this prediction on knowledge 
of wolf and woodland caribou population dynamics 
in Canada and my assessment that secure winter 
escape habitat is limited on Isle Royale. The 
prediction's accuracy might be improved with 
additional population and habitat suitability model- 
ing or with an experimental release. 

The questions posed by caribou restoration are not 
unique; wolves may never be a self-sustaining 
species on Isle Royale either, being unable to reach 
a 100-year or longer minimum viable population. 
The challenge is to determine how important these 
large animals are to the island ecosystem and how 
important the island is to regional population 
viability of these species. The answers may deter- 
mine how aggressively species are managed. Wright 
et al. (1933) anticipated that "not one park is large 
enough to provide year-round sanctuary for ad- 
equate populations of all resident species ... [or to] 
guarantee against the invasion of external influ- 
ences." While they promoted restoration of species 
to achieve an "original, natural state" in national 



parks, they also cautioned that "due care must be 
taken that management does not create an even 
more artificial condition in place of the one it would 
correct" (ibid). 

Reflecting recent trends in biodiversity conserva- 
tion, Soule (1986:7-8) presents a case for aggres- 
sive management of large, rare species: 

Viability, in the strict sense, will be impossible for 
some populations in certain situations, but wise 
management may still be able to insure the per- 
sistence of a relatively 'natural' community or sys- 
tem. Probably the majority of nature reserves in the 
world are too small to contain more than a few fam- 
ily groups of primates or herds of large ungulates. 
Such tiny groups in vest pocket reserves may con- 
tribute virtually nothing to the viability of the spe- 
cies as a whole, but their survival may be impor- 
tant, nonetheless, for ecological and social rea- 
sons. . . . When such groups die out, therefore, they 
should simply be replaced from whatever source is 
available. 

If maintaining regional biodiversity is a goal, 
management of isolated nature reserves and national 
parks will increasingly require coordination outside 
reserve and park boundaries. Perhaps more than 
any other mammal on Isle Royale, woodland 
caribou were historically part of a regional meta- 
population. Thus, the value of restoring caribou on 
Isle Royale should be assessed not only from the 
park's perspective but also as part of a regional 
effort to preserve the subspecies at the southcentral 
boundary of its range. 

Translocating woodland caribou from the Slate 
Islands to Isle Royale and other regional sites — if 
practical — would disperse the genetic stock of Lake 
Superior caribou and help to reduce the chances of 
a catastrophic loss of this ecologically and geneti- 
cally unique subgroup. But sustaining these small 
herds, isolated from each other by human-altered 
landscapes, would require periodic intervention. Isle 
Royale, a national park and designated wilderness 
supporting unique floral and faunal communities, 
may not be an appropriate setting for this type of 
biodiversity management. Based on the NPS 



40 



Conclusions 



management policy that restored populations be 
"self-perpetuating," Isle Royale would not be an 
appropriate site for caribou restoration under 
current conditions. 



Conclusions 41 



Literature Cited 



Abraham, K., B. Darby, Q. Day, B. Foster, J. McNicol, 
G. Racey, and T. Timmermann. 1990. Timber 
management guidelines for the provision of wood- 
land caribou habitat. Draft report. Ontario Ministry 
of Natural Resources, Thunder Bay, Ontario. 32 pp. 

Adams, C.C. 1909. An ecological survey of Isle Royale, 
Lake Superior. Michigan State Biological Survey 
Report. Wynkoop Hallenbeck Crawford Co., Lan- 
sing, Michigan. 468 pp. 

Ahti, T. and R.L. Hepburn. 1967. Preliminary studies 
on woodland caribou range, especially on lichen 
stands, in Ontario. Ontario Department of Lands and 
Forest Resources, Report (Wildlife) No. 74. 134 pp. 

Anderson, R. 1988. Interview with T. Cochrane, 16 
October 1988 at Grand Portage, Minnesota. Isle 
Royale National Park Oral History No. 68, Mott 
Island. 

Anderson, R.C. 1971. Neurological disease in reindeer 
(Rangifer tarandus tarandus) introduced into 
Ontario. Canadian Journal of Zoology 49:159-166. 

Anderson, R.C. and A.K. Prestwood. 1981. Lung- 
worms. Pages 266-317 in W.R. Davidson, FA. 
Hayes, V.F Nettles, and FE. Kellog, editors. 
Diseases and Parasites of White-tailed Deer. Miscel- 
laneous Publication 7. Tall Timbers Research 
Station, Tallahassee, Florida. 

Anderson, R.C. and U.R. Strelive. 1968. The experi- 
mental transmission of Pneumostrongylus tenuis to 
caribou (Rangifer tarandus terranovae). Canadian 
Journal of Zoology 46:503-510. 

Anonymous. 1928. Many moose and beaver found on 
Isle Royale. Detroit News, 20 August 1928. Copy in 
Detroit News collection, Michigan Historical 
Collections, Ann Arbor. 



Bergerud, A.T. 1972. Food habits of Newfoundland 
caribou. Journal of Wildlife Management 36(3):913- 
923. 

Bergerud, A.T. 1974a. Decline of caribou in North 
America following settlement. Journal of Wildlife 
Management 38(4):757-770. 

Bergerud, A.T. 19746. The role of the environment in 
the aggregation, movement and disturbance behav- 
iour of caribou. Pages 552-584 in V. Geist and F 
Walther, editors. The Behaviour of Ungulates and Its 
Relation to Management, Vol. 2. IUCN New Series 
Publication 24, Morges, Switzerland. 

Bergerud, A.T. 1978. Caribou. Pages 83-101 in J.L. 
Schmidt and D.L. Gilbert, editors. Big Game of 
North America: Ecology and Management. Stack- 
pole Books, Harrisburg, Pennsylvania. 

Bergerud, A.T. 1980. A review of the population 
dynamics of caribou and wild reindeer in North 
America. Pages 556-581 in E. Reimers, E. Gaare, 
and S. Skenneberg, editors. Proceedings of the 
Second International Reindeer/Caribou Symposium, 
Direktoratet for vilt og frskvannsfisk, Tronheim, 
Norway. 

Bergerud, A.T. 1983. The natural population control of 
caribou. Pages 14-61 in FL. Bunnell, D.S. Eastman, 
and J.M. Peek, editors. Symposium on Natural 
Regulation of Wildlife Populations. (1978). Forest, 
Wildlife and Range Experiment Station, University 
of Idaho, Moscow. 

Bergerud, A.T. 1985. Antipredator strategies of caribou: 
dispersion along shorelines. Canadian Journal of 
Zoology 63(6): 1324- 1329. 

Bergerud, A.T. 1988. Caribou, wolves and man. Trends 
in Ecology and Evolution 3:68-72. 



Baker, R.H. 1983. Michigan Mammals. Michigan State 
University Press, East Lansing. 

Bergerud, A.T. 1971. The population dynamics of 
Newfoundland caribou. Wildlife Monograph 25. 55 
pp. 



43 



Bergerud, A.T. 1989a. The abundance, distribution and 
behaviour of caribou in Pukaskwa National Park, 
1972-1988. Unpublished manuscript, Bergerud and 
Associates Contract Report No. 88-CPS-PUK, 
Scientific Review of Caribou Management Activities 
in Pukaskwa National Park, Fulford Harbour, British 
Columbia. 57+ pp. 

Bergerud, A.T. 1989b. Aerial census of caribou and 
wolves in Wabakimi Provincial Park. Unpublished 
manuscript, Bergerud and Associates, Fulford 
Harbor, British Columbia. 41 pp. 

Bergerud, A.T, H.E. Butler, and D.R. Miller. 1984. 
Anti-predator tactics of calving caribou: dispersion 
in mountains. Canadian Journal of Zoology 62:15- 
66-1575. 

Bergerud, A.T. and J. P. Elliot. 1986. Dynamics of 
caribou and wolves in northern British Columbia. 
Canadian Journal of Zoology 64:1515-1529. 

Bergerud, A.T. and R.E. Page. 1987. Displacement and 
dispersion of parturient caribou at calving as 
antipredator tactics. Canadian Journal of Zoology 
65:1597-1606. 

Bergerud, A.T, R. Ferguson, and H.E. Butler. 1990. 
Spring migration and dispersion of woodland 
caribou at calving. Animal Behavior 39(2):360-368. 



Brown, W.K., J. Hout, P. Lamothe, S. Luttich, M. Pare, 
G. St. Martin, and J.B. Theberge. 1986. The distri- 
bution and movement patterns of four woodland 
caribou herds in Quebec and Labrador. Pages 43-49 
in A. Gun, FL. Miller, and S. Skjenneberg, editors. 
Proceedings of the Fourth International Reindeer/ 
Caribou Symposium, Rangifer Special Issue No. 
1:43-49. 

Brown, W.K. and J.B. Theberge. 1990. The effect of 
extreme snowcover on feeding-site selection by 
woodland caribou. Journal of Wildlife Management 
54(1):161-168. 

Cleland, C.E. 1966. The prehistoric animal ecology and 
ethnohistory of the Upper Great Lakes region. 
University Michigan Museum Anthropology, 
Anthropological Paper No. 29., Ann Arbor, Michi- 
gan. 294 pp. 

Cleland, C.E. 1968. Analysis of the fauna of the Indian 
Point site on Isle Royale in Lake Superior. Michigan 
Archeology 14:143-146. 

Commission on Research and Resource Management in 
the National Park System. 1989. National parks: 
from vignettes to a global view. Report from the 
Commission on Research and Resource Management 
in the National Park System. National Parks and 
Conservation Association, Washington, D.C. 13 pp. 



Bergerud, A.T. and W.E. Mercer. 1989. Caribou 
introductions in eastern North America. Wildlife 
Society Bulletin 17:111-120. 

Blackwell, B., N. Deschampe, Jr., L. Thibeault, T 
Olson, and G. Caribou. 1983. A History of Kitchi 
Onigaming: Grand Portage and Its People. Minne- 
sota Chippewa Tribe, Cass Lake, Minnesota. 



Conary, Mrs. H. 1939. Interview titled, "Early life on 
Isle Royale," published in the Daily Mining Gazette, 
Houghton, Michigan, 21 June 1939. Copy in Isle 
Royale National Park Library, Mott Island. 

Crane, H.R. and J.R. Griffin. 1965. University of 
Michigan Radiocarbon Dates X, Vol. 7, pages 126- 
128. New Haven, Connecticut. 



Blair, E.H. 1911. The Indian tribes of the upper 

Mississippi Valley and region of the Great Lakes as 
described by Nicholas Perrot, Vol. 1 . Clark, Cleve- 
land. 372 pp. 

Breckenridge, W.J. 1949. A century of Minnesota 
wildlife. Minnesota History 30:123-134. 



Cringan, A.T. 1956. Some aspects of the biology of 
caribou and a study of the woodland caribou range of 
the Slate Islands, Lake Superior, Ontario. Unpub- 
lished MA. thesis, University of Toronto, Ontario. 
300 pp. 

Cringan, A.T. 1957. History, food habits and range 
requirements of the woodland caribou of continental 
North America. Transactions of the North American 
Wildlife and Natural Resources Conference 22:485- 
501. 



44 



Literature Cited 



Cumming, H.G. and D.B. Beange. 1987. Dispersion 
and movements of woodland caribou near Lake 
Nipigon, Ontario. Journal of Wildlife Management 
51:69-78. 

Cutler, M.R. 1989. Letter to James Ridenour, director, 
U.S. Department of the Interior, National Park 
Service. Copy in Isle Royale National Park files, 
Houghton, Michigan. 



Edmonds, E.J. and M. Bloomfield. 1984. A study of 
woodland caribou (Rangifer tarandus caribou) in 
west central Alberta 1979-1983. Unpublished report, 
Alberta Department of Energy and Natural Resourc- 
es, Fish and Wildlife Division, Edmonton, Alberta. 
203 pp. 

Edwards, J. 1983. Diet shifts in moose due to predator 
avoidance. Oecologia 60:185-189. 



Darby, W.R. 1979. Seasonal movements, habitat 
utilization and population ecology of woodland 
caribou (Rangifer tarandus caribou Gemlin) in the 
Wallace- Aikens Lake region of south-eastern 
Manitoba. Unpublished M.S. thesis, University of 
Manitoba, Winnipeg. 187 pp. 

Darby, W.R. and W.O. Pruitt, Jr. 1984. Habitat use, 
movements and grouping behavior of woodland 
caribou, Rangifer tarandus caribou, in southeastern 
Manitoba. Canadian Field Naturalist 98(2): 184-190. 

Darby, W.R., H.R. Timmermann, J.B. Snider, K.F. 
Abraham, R.A. Stefanski, and C.A. Johnson. 1989. 
Woodland caribou in Ontario: background to policy. 
Unpublished report, Ontario Ministry of Natural 
Resources, Toronto. 38 pp. 



Erskine, L.Y. 1922. Isle Royale invasion threatens 
kingdom of the moose. Detroit News, 17 December 
1922. 

Euler, D.L., B. Snyder, and H.R. Timmerman. 1976. 
Woodland caribou and plant communities on the 
Slate Islands, Lake Superior. Canadian Field 
Naturalist 90(1): 17-21. 

Fashingbauer, B.A. 1965. The woodland caribou in 
Minnesota. Pages 133-166 in J.B. Moyle, editor. Big 
Game in Minnesota. Minnesota Department of 
Conservation Technical Bulletin 9, St. Paul. 

Ferguson, S.H. 1983. Why are caribou on Pic Island? 
Unpublished M.S. thesis, University of Victoria, 
British Columbia. 180 pp. 



de Vos, A. and R.L. Peterson. 1951. A review of the 
status of the woodland caribou (Rangifer caribou) in 
Ontario. Journal of Mammalogy 32(3):329-337. 

Dustin, F. 1946. Isle Royale place names. Michigan 
History 30:681-722. 

Edisen, P. n.d. Interview with anonymous. Isle Royale 
National Park Oral History No. 3, Mott Island. 



Ferguson, S.H., A.T Bergerud, and R. Ferguson. 1988. 
Predation risk and habitat selection in the persis- 
tence of a remnant caribou population. Oecologia 
76:236-245. 

Ferguson, W.PF 1922a. Letter to Albert Stoll, Detroit, 
Michigan, 25 January 1922. Copy in Detroit News 
collection, Michigan Historical Collections, Ann 
Arbor. 



Edisen, P. 1975. Interview with J. Lund, 12 August 
1975, at Edisen Fishery, Isle Royale, Michigan. Isle 
Royale National Park Oral History No. 64, Mott 
Island. 

Edisen, P. and L. Edisen. 1965. Interview with L. 

Rakestraw, 3 September 1965, at Edisen Fishery, Isle 
Royale, Michigan. Isle Royale National Park Oral 
History No. 22, Mott Island. 

Edmonds, E.J. 1988. Population status, distribution, 
and movements of woodland caribou in west central 
Alberta. Canadian Journal of Zoology 66:817-826. 



Ferguson, W.PF 19226. On Isle Royale with the Isle 
Royalists. Detroit News, 5 March 1922. Copy at Isle 
Royale National Park Library, Mott Island. 

Ferguson, W.PF 1922c. Wildlife and adventures along 
the Great Desor Trail on Isle Royale. Detroit News, 
28 May 1922. 

Ferguson, W.PF 1926. Letter to the editor. Michigan 
History 10:471-473. 

Foster, J.W 1851. Letter to Spencer F. Baird. Copy in 
Isle Royale National Park Library, Mott Island. 



Literature Cited 



45 



Foster, J.W. and J.D. Whitney. 1850. Report on the 
geology and topography of a portion of the Lake 
Superior Land District, Part 1 : Copper Lands. Pages 
179-192 in M.E. Gilcrist, editor. Isle Royale Survey. 
Inland Seas, Washington, D.C. (Fall 1968). 



Gogan, P.J.P., P.A. Jordan, and J.L. Nelson. 1990. 
Planning to reintroduce woodland caribou to 
Minnesota. Transactions of the North American 
Wildlife and Natural Resources Conference 55:599- 
608. 



Foster, WH. 1917. Review of wild life— Isle Roy- 
ale-1916-1917. Michigan Sportsman 4:16-18. 

Fruetel, M. and M.W. Lankester. 1989. Gastrointestinal 
helminths of woodland and barren ground caribou 
(Rangifer tarrandus) in Canada, with keys to 
species. Canadian Journal of Zoology 67:2253-2269. 

Fuller, T.K. and L.B. Keith. 1981. Woodland caribou 
population dynamics in northeastern Alberta. 
Journal of Wildlife Management 45(1): 197-211. 

Gillman, H. 1873. The caribou on Lake Superior. 
American Naturalist 7:751. 

Gilpin, M.E. 1987. Spatial structure and population 
vulnerability. Pages 125-139 in M.E. Soule, editor. 
Viable Populations for Conservation. Cambridge 
University Press, England. 



Goodwin, L. 1990. Woodland caribou in northwestern 
Ontario: why they are different. Ontario Ministry of 
Natural Resources, Northwestern Ontario Forest 
Technology Development Unit, Thunder Bay, Report 
TN-07. 7 pp. 

Grier, J. 1980. Ecology: a simulation model for small 
populations of animals. Creative Computing 6:116- 
121. 

Griffith, B., J.M. Scott, J.W. Carpenter, and C. Reed. 
1989. Translocation as a species management tool: 
status and strategy. Science 245:477-480. 

Hansen, H.L., L.W. Krefting, and V. Kurmis. 1973. The 
forest of Isle Royale in relation to fire history and 
wildlife. University of Minnesota Agricultural 
Experiment Station Technical Bulletin 294, Forestry 
Series 13, St. Paul, Minnesota. 43 pp. 



Gilpin, M.E. and M.E. Soule. 1986. Minimum viable 
populations: processes of species extinction. Pages 
19-34 in M.E. Soule, editor. Conservation Biology: 
The Science of Scarcity and Diversity. Sinauer 
Associates, Sunderland, Massachusetts. 



Hickerson, H. 1974. Ethnohistory of Chippewa of Lake 
Superior. American Indian Ethnohistory: North 
Central and Northeastern Indians, Vol. 3. U.S. 
Indian Claims Commission. Garland Press, New 
York. 295 pp. 



Glines, M.V and W.M. Samuels. 1989. Effect of 
Dermacentor albipictus (Acari: Ixodidae) on blood 
composition, weight gain and hair coat of moose, 
Alces alces. Experimental & Applied Acarology 
6:197-213. 

Gogan, P.J.P. 1990. Considerations in the reintro- 
duction of native mammalian species to restore 
natural ecosystems. Natural Areas Journal 10:210- 
217. 

Gogan, P.J.P. and J.F Cochrane. 1994. Restoration of 
woodland caribou to the Lake Superior region. Pages 
219-242 in M.L. Bowles and C.J. Whelan, editors. 
Restoration of Endangered Species: conceptual 
issues, planning and implementation. Cambridge 
University Press, Cambridge, England. 394pp. 



Hickie, P.F n.d. (1937?). Michigan moose. Michigan 
Department of Conservation, Game Division, 
Lansing. 57 pp. 

Holte, I. 1980. Interview with T Cochrane, 17 August 
1980, Wright Island, Isle Royale, Michigan. North- 
east Minnesota History Center, University of 
Minnesota, Duluth. 

Holte, I. and E. Holte. 1965. Interview with L. 

Rakestraw, 10 September 1965, Wright Island, Isle 
Royale, Michigan. Isle Royale National Park Oral 
History No. 2. 

Houghton, J. 1892. Letter to the chairman and directors 
of the Wendigo Coppe Co. Ltd., 10 October 1982. 
Papers of the Isle Royale Land Corporation and the 
Wendigo Copper Mining Company, Isle Royale 
National Park Library, Mott Island. 



46 



Literature Cited 



Jackson, C.T. 1852. Report to Congress. U.S. Senate 
Executive Document No. 112, 32nd Congress, Serial 
622, Washington, D.C. 

Jakimchuk, R.D., S.H. Ferguson, and L.G. Sopuck. 
1987. Differential habitat use and sexual segregation 
in the central arctic caribou herd. Canadian Journal 
of Zoology 65:534-541. 

James, E., editor. 1956 (1830). A Narrative of the 
Captivity and Adventures of John Tanner During 
Thirty Years Residence Among the Indians in the 
Interior of North America. Ross and Haines, 
Minneapolis. 427 pp. 

Janke, R.A., D. McKaig, and R. Raymond. 1978. 
Comparison of presettlement and modern upland 
boreal forests on Isle Royale National Park. Forest 
Science 24(1):1 15-121 . 

Jeffery, W. 1942. Interview notes titled "Early life at 
Island Mine" 1873-1874. Interview with anonymous, 
February 1942, Isle Royale National Park Library, 
Mott Island. 

Johnson, Mike. 1942. Interview with H.E. Bailey, 1 
October 1942. Notes from oral history, Isle Royale 
National Park Library, Mott Island. 

Johnson, Milford. 1965. Interview with L. Rakestraw. 
Isle Royale National Park. Oral History No. 5, Mott 
Island. 

Jordan, PA. 1978. Miscellaneous information and 
speculation on moose and their impact on vegetation 
and on some other herbivores at Isle Royale. Un- 
published report, Department of Entomology, 
Fisheries and Wildlife, University of Minnesota, St. 
Paul. 13 pp. 

Jordan, PA. 1981. Wildlife of Isle Royale. Isle Royale 
Natural History Association, Houghton, Michigan. 
43 pp. 

Kapler, C.J. 1904. Indian Affairs: Laws and Treaties. 
U.S. Government Printing Office, Washington, D.C. 

Karamanski, T.J., R. Zeitlin, and J. DeRose. 1988. 
Narrative history of Isle Royale National Park. Mid- 
America Research Center, Loyola University, 
Chicago, Illinois. 386 pp. 



Karns, P.D. and PA. Jordan. 1969. Pneumostrongylus 
tenuis in moose on a deer-free island. Journal of 
Wildlife Management 33(2) :43 1-433. 

Karns, P.D. and E.L. Lindquist. 1986. Reintroduction 
of the woodland caribou to the Superior National 
Forest. Unpublished report, Minnesota Department 
of Natural Resources, St. Paul. 58 pp. 

Klein, D.R. 1968. The introduction, increase, and crash 
of reindeer on St. Matthew Island. Journal of 
Wildlife Management 32(2):350-367. 

Krefting, L.W 1969. The rise and fall of the coyote on 
Isle Royale. Michigan Naturalist 20(4):24-31. 

Krefting, L.W. 1974. The ecology of the Isle Royale 
moose with special reference to the habitat. Univer- 
sity of Minnesota Agricultural Experiment Station 
Technical Bulletin No. 297 (Forest Series No. 15), 
St. Paul. 75 pp. 

Krefting, L.W., H.L. Hansen, and M.P Meyer. 1970. 
Vegetation type map of Isle Royale National Park. 
Published by the U.S. Bureau of Sport Fisheries and 
Wildlife, Denver, Colorado. Copy in Isle Royale 
National Park Library, Mott Island. 

Lacy, R. 1990. VORTEX: simulation model of sto- 
chastic population change. Version 8.0. Unpublished 
manuscript, Chicago Zoological Park, Brookfield, 
Illinois. 4 pp. 

La More, W.H. 1923. Interview published in the 
Milwaukee Journal, 4 February 1923. Copy in Isle 
Royale National Park Library, Mott Island. 

Lande, R. and G.F Barrowclough. 1987. Effective 
population size, genetic variation and their use in 
population management. Pages 85-123 in M.E. 
Soule, editor. Viable Populations for Conservation. 
Cambridge University Press, England. 

Lankester, M.W. 1976. A Protostrongylid nematode of 
woodland caribou and implications in moose 
management. North American Moose Conference 
and Workshop 12:173-190. 



Literature Cited 



47 



Lankester, M.W. and D. Fong. 1989. Distribution of 
Elaphostrongyline nematodes (Metestrongyloidea:- 
protostrongylidae) in cervidae and possible effects of 
moving Rangifer species into and within North 
America. Alces 25:133-145. 

Lankester, M.W. and P.L. Hauta. 1989. Parelaphost- 
rongylus andersoni (Nematoda:Protostrongylidae) in 
caribou (Rangifer tarandus) of northern and central 
Canada. Canadian Journal of Zoology 67:1966- 
1975. 

Lankester, M.W. and T.H. Northcott. 1979. Elapho- 
strohgylus cervi Cameron 1931 (Nematoda: Metast- 
rongyloidea) in caribou (Rangifer tarandus caribou) 
in Newfoundland. Canadian Journal of Zoology 
57:1384-1392. 

Leopold, A.S., S. Cain, C. Cottam, I. Gabrielson, and T. 
Kimball. 1963. Wildlife management in the national 
parks. Transactions of the North American Wildlife 
and Natural Resources Conference 28:29-44. 



Master, J. 1913. Interview notes titled, "Recollections of 
an old Copper Country resident," interview with 
anonymous, 17 March 1913. Isle Royale National 
Park Library, Mott Island. 

McCollough, M. 1987. A management and research 
plan for the reintroduction of woodland caribou to 
Maine. Unpublished report, Maine Caribou Project, 
University of Maine, Orono. 43 pp. 

McCollough, M. and B. Connery. 1990. An evaluation 
of the Maine caribou reintroduction project, 1986 to 
1989. Unpublished report, Maine Caribou Project, 
University of Maine, Orono. 41+ pp. 

Mech, L.D. 1966. The wolves of Isle Royale. U.S. 
Department of the Interior, National Park Service 
Fauna Series No. 7, Washington, D.C. 210 pp. 

Mech, L.D., M.E. Nelson, and H.F. Drabik. 1982. 
Reoccurence of caribou in Minnesota. American 
Midland Naturalist 108(l):206-208. 



Linn, R.M. 1957. The spruce-fir, maple-birch transition 
in Isle Royale National Park, Lake Superior. Unpub- 
lished Ph.D. dissertation, Duke University, Raleigh, 
North Carolina. 98 pp. 

Lytwyn, V.P. 1986. The Fur Trade of the Little North: 
Indians, Peddlers, and Englishmen East of Lake 
Winnipeg, 1760-1821. Rupert's Land Research 
Center, University of Winnipeg, Manitoba. 208 pp. 

Martin, C. 1988. The history and reestablishment 
potential of marten, lynx, and woodland caribou on 
Isle Royale. Isle Royale National Park Resource 
Management Report 88-2, Houghton, Michigan. 68 
pp. 

Martin, T.J. 1989. Animal remains from archaeological 
sites at Isle Royale National Park, Michigan: 1988 
season. Quaternary Studies Project Technical Report 
89-512-30. Illinois State Museum Society, Spring- 
field. 

Martin, T.J. and M.C. Maulis. 1989. Animal remains 
from archaeological sites at Isle Royale National 
Park: 1987 season. Quaternary Studies Project 
Technical Report 89-389-8. Illinois State Museum 
Society, Springfield. 



Merritt, G. n.d. Interview with anonymous. Isle Royale 
National Park Oral History No. 20, Mott Island. 

Merritt, G. 1974. Interview with G. Merritt, September 
1974. Isle Royale National Park Oral History No. 12, 
Mott Island. 

Miller, FL. 1982. Caribou. Pages 923-959 in J. A. 
Chapman and G.A. Feldhamer, editors. Wild 
Mammals of North America: Biology, Management, 
and Economics. The Johns Hopkins University 
Press, Baltimore. 

Miller, G.S., Jr. 1897. Notes on the mammals of 
Ontario. Proceedings of the Boston Society of 
Natural History 28:1-44. 

Miller, V. 1986. Interview with T Cochrane, 12 
October 1986, Ahmeek, Michigan. Isle Royale 
National Park Oral History No. 66, Mott Island. 

Murie, A. 1934. The moose of Isle Royale. University of 
Michigan Museum Zoology Miscellaneous Publica- 
tion No. 25:1-44. 

Myers, A. 1851. Journey to Siskowit Mine in 1851. 
Manuscript at The Historical Society of Pennsyl- 
vania, Philadelphia. Copy at Isle Royale National 
Park Library, Mott Island. 



48 



Literature Cited 



Oastler, F. 1929. A survey of Isle Royale-Lake Superior. 
Survey for the National Park Service. Copy in 
Detroit News collection, Michigan Historical 
Collections, Ann Arbor. 

Oikarinen, P. 1979. Island Folk: The People of Isle 
Royale. Isle Royale Natural History Association, 
Houghton, Michigan. 158 pp. 

Ontario Ministry of Natural Resources. 1986. Slate 
Islands Provincial Park Management Plan; prelim- 
inary draft. Unpublished report, Ontario Ministry of 
Natural Resources, Terrace Bay. 27 pp. 

Peterson, R.O. 1977. Management implications of wolf- 
moose research, Isle Royale National Park, Michi- 
gan. Unpublished report, Isle Royale National Park, 
Houghton. 14 pp. 

Peterson, R.O. 1990. Ecological studies of wolves on 
Isle Royale; Annual Report 1989-1990. Isle Royale 
Natural History Association, Houghton, Michigan. 
14 pp. 

Peterson, R.O. 1992. Ecological studies of wolves on 
Isle Royale; Annual Report 1991-1992. Isle Royale 
Natural History Association, Houghton, Michigan. 
16 pp. 

Peterson, R.O. and D.L. Allen. 1974. Snow conditions 
as a parameter in moose-wolf relationships. Le 
Naturaliste Canadien 101(3,4):48 1-492. 

Peterson, R.O. and J.M. Thurber. 1990. Study of causes 
of wolf decline, Isle Royale National Park. Unpub- 
lished report, Isle Royale National Park, Houghton. 
32+ pp. 

Phillips, B. n.d. (1940s). Notes from interview with 
anonymous, Isle Royale National Park Library, Mott 
Island. 

Pybus, M.J. and W.M. Samuel. 1984. Parelapho- 
strongylus andersoni (Nematoda: Protostrongylidae) 
and P. odocolei in two cervid definitive hosts. 
Journal of Parasitology 70:507-515. 

Quaife, M.M., editor. 1921. Alexander Henry's travel 
and adventures in the years 1769-1776. Lakeside 
Press, Chicago. 340 pp. 



Rakestraw, L. 1965. Historic Mining of Isle Royale. Isle 
Royale Natural History Association, Houghton, 
Michigan. 20 pp. 

Raymond, C.L. n.d. In Memory of Our Fishing Trip in 
the Lake Superior Region, Summer of 1897. 
McClurg & Company, Chicago. 34 pp. 

Riis, PB. 1938. Woodland caribou and time. Parks and 
Recreation 21:529-535, 594-600, 639-645; 22:23- 
30. 

Risenhover, K.L. and S.A. Maass. 1987. The influence 
of moose on the composition and structure of Isle 
Royale forests. Canadian Journal of Forest Resources 
17:357-364. 

Samuel, W.M. 1989. Locations of moose in north- 
western Canada with hair loss probably caused by 
the winter tick, Dermacentor albipictus (Acari: 
Ixodidae). Journal of Wildlife Diseases 25(3):436- 
439. 

Schaeffer, J. A. and W.O. Pruitt, Jr. 1991. Fire and 
woodland caribou in southeastern Manitoba. Wildlife 
Monograph 116. 39 pp. 

Schorger, A.W. 1942. Extinct and endangered mam- 
mals and birds of the Upper Great Lakes region. 
Transactions of the Wisconsin Academy of Science, 
Arts, and Letters 34:23-44. 

Scott, E.H. 1924. Letter to Albert Stoll, Jr., Detroit, 
Michigan, 7 March 1924. Copy in Detroit News 
collection, Michigan Historical Collections, Ann 
Arbor. 

Scott, W.P. 1925. Reminiscences of Isle Royale. 
Michigan History 9:398-412. 

Servheen, G. 1988. Selkirk Mountains caribou trans- 
plant: October 1987 - September 1988. Unpublished 
report, Idaho Department of Fish and Game, Boise. 
36 pp. 

Servheen, G. 1989. Selkirk Mountains caribou trans- 
plant: October 1988 - September 1989. Draft report, 
Idaho Department of Fish and Game, Boise. 45+ pp. 

Shaffer, M.L. 1981. Minimum population sizes for 
species conservation. BioScience 31:131-134. 



Literature Cited 



49 



Shoesmith, M.W. and D.R. Storey. 1977. Movements 
and associated behaviour of woodland caribou in 
central Manitoba. Proceedings of the International 
Congress of Game Biologists 13:51-64. 

Simkin, D.W. 1965. A preliminary report of the 

woodland caribou study in Ontario. Ontario Depart- 
ment of Lands and Forests Report (Wildlife) No. 59, 
Toronto. 76 pp. 

Skadberg, J. 1987. Interview with D. Snyder and R. 
Peterson, January 1987, at Grand Marais, Minne- 
sota. Isle Royale National Park Oral History No. 60, 
Mott Island. 



Stoll, A., Jr., editor. 1924a. Biennial report, 1921-1922. 
Michigan Department of Conservation, Lansing. 358 
pp. 

Stoll, A., Jr. 19246. Letter to Honorable Hubert Work, 
Washington, D.C., 13 February 1924. Copy in 
Detroit News collection, Michigan Historical 
Collections, Ann Arbor. 

Stoll, A., Jr. 1926. An unspoiled and little known 
wonderland of the north. American Forests and 
Forest Life, August 1926. Typed manuscript in 
Detroit News collection, Michigan Historical 
Collections, Ann Arbor. 



Skoog, R.O. 1968. Ecology of the caribou in Alaska. 
Ph.D. dissertation. University of California, Berke- 
ley. 699 pp. 

Slavik, A.D. and R.A. Janke. 1987. The vascular flora 
of Isle Royale National Park. Michigan Botanist 
26(3):91-134. 

Snyder, J.D. and R.A. Janke. 1976. Impact of moose 
browsing on boreal-type forests of Isle Royale 
National Park. American Midland Naturalist 
95(l):79-92. 

Soule, M.E. 1986. Conservation biology and the real 
world. Pages 1-12 in M.E. Soule, editor. Conser- 
vation Biology: The Science of Scarcity and 
Diversity. Sinauer Associates, Sunderland, Massa- 
chusetts. 

Soule, M.E. 1987. Introduction. Pages 1-10 in M.E. 
Soule, editor. Viable Populations for Conservation. 
Cambridge University Press, England. 

Soule, M.E. and Simberloff. 1986. What do genetics 
and ecology tell us about the design of nature 
reserves. Biological Conservation 35:19-40. 



Strommen, N.D. 1969. Isle Royale National Park, 
Michigan. Climatic summaries of resort areas. U.S. 
Department of Commerce, Environmental Science 
Services Administration. Climatography of the U.S. 
No. 21-20-1. 4 pp. 

Summerfield, B. 1985a. Environmental assessment, 
Selkirk Mountains caribou herd augmentation. U.S. 
Department of Agriculture, Forest Service, Idaho 
Panhandle National Forests, Coeur d'Alene. 13+ pp. 

Summerfield, B. 19856. Selkirk Mountains caribou 
herd augmentation: a cooperative interagency plan. 
Draft report, U.S. Department of Agriculture, Forest 
Service, Idaho Panhandle National Forests, Coeur 
d'Alene, Idaho. 18+ pp. 

Swanson, E.B. 1940. The use and conservation of 
Minnesota game, 1850-1900. Unpublished Ph.D. 
dissertation, University of Minnesota, Minneapolis. 
194 pp. 

Swanson, G., T Surber, and T.S. Roberts. 1945. The 
mammals of Minnesota. Minnesota Department of 
Conservation Technical Bulletin No. 2, St. Paul. 107 
pp. 



Stardom, R.R.P. 1975. Woodland caribou and snow 
conditions in southeast Manitoba. Pages 436-461 in 
J.R. Luick, PC. Lent, D.R. Klein, and R.G. White, 
editors. Proceedings of the First International Rein- 
deer/Caribou Symposium, University of Alaska 
Biological Papers, Special Report No. 1. 

Stephens, PW. and R.O. Peterson. 1984. Wolf-avoid- 
ance strategies of moose. Holarctic Ecology 7:239- 
244. 



Thomas, D.C., S.J. Barry, and H.P Kiliaan. 1989. Fetal 
sex ratios in caribou: maternal age and condition 
effects. Journal of Wildlife Management 53(4):885- 
890. 

Timm, R.M. 1975. Distribution, natural history, and 
parasites of mammals of Cook County, Minnesota. 
University Minnesota Bell Museum Natural History 
Occasional Paper 14. 56 pp. 



so 



Literature Cited 



U.S. Department of the Interior, National Park Service. 
1988. NPS management policies. U.S. Department 
of the Interior, National Park Service, Washington, 
D.C. 

U.S. Indian Service. 1840. Annual Report for 1840. 
Washington, D.C. 

Warren, F.M. 1924. Letter to Albert Stoll, Jr., Detroit, 
Michigan, 22 March 1924. Copy in Detroit News 
collection, Michigan Historical Collections, Ann 
Arbor. 



Wright, G.M., J.S. Dixon, and B.H. Thompson. 1933. 
Fauna of the National Parks of the United States: a 
preliminary survey of faunal relations in National 
Parks. U.S. National Park Service Fauna Series No. 
1, July 1934, Washington, D.C. 157 pp. 

Wright, G.M. and B.H. Thompson. 1935. Fauna of the 
National Parks of the United States: Wildlife 
Management in the National Parks. U.S. National 
Park Service Fauna Series No. 2 — July 1934, 
Washington, D.C. 142 pp. 



Warren, F.M. 1926. The Wildlife of Isle Royale. 
American Game 15(1): 15- 17. 

Warren, F.M. 1929. Letter to Albert Stoll, Jr., Detroit, 
Michigan, 24 October 1929. Copy in Detroit News 
collection, Michigan Historical Collections, Ann 
Arbor. 

Warren, L. 1834. Letter to Ramsey Crooks, New York, 
16 October 1834. Copy in American Fur Company 
Papers, New York Historical Society, New York. 

Wayne, R.K., N. Lehman, D. Girman, P.J.P. Gogan, 
D.A. Gilbert, K. Hansen, R.O. Peterson, U.S. Seal, 
A. Eisenhower, L.D. Mech, and R.J. Krumenaker. 
1991. Conservation genetics of the endangered Isle 
Royale gray wolf. Conservation Biology 5(1):41-51. 

Welch, D.A., W.M. Samuels, and C.J. Wilke. 1990. 
Dermacentor albipictus (Acari, Ixodidae) on captive 
reindeer and free-ranging woodland caribou. Journal 
of Wildlife Diseases 26(3):410-411. 

Wepruk, R. 1986. Caribou investigations of Woodland 
Caribou Park, 1985. Woodland Caribou Park 
Technical Report No. 3, Ontario Ministry of Natural 
Resources, Red Lake. 

West, G.A. 1929. Copper: its mining and use by the 
aborigines of the Lake Superior region. Report of the 
McDonald-Massee Isle Royale Expedition 1928. 
Bulletin Milwaukee Public Museum, Vol. X. 185 pp. 

Willebrands, B.O. 1923. Letter to Honorable Governor 
Grosbeck, Lansing, Michigan, 5 September 1923. 
Copy in Detroit News collection, Michigan Histori- 
cal Collections, Ann Arbor. 



Literature Cited 



51 



Postscript 



"I saw caribou when I was young. One 
spring, we came down on the steamer 
America, oh, it was in April, about the 20th 
of April, and there was still a lot of ice in 
Rock Harbor channel. And you know, those 
animals, there' d usually be a lot of them 
together. There 'd be maybe seven, eight or 
more. And then this one time across from 
Anderson's place, up, right across from 
Caribou Island, there were 10 or 12 of 'em 
walking along the ice. It was still strong 
enough to walk on, for caribou. And since 
then, I only seen occasional, you know, not 
very many. But finally they disappeared 
altogether. The moose came in, they seemed 
to leave." (Milford Johnson, interviewed by 
Lawrence Rakestraw, 1965; Isle Royale 
Oral History No. 5.) 

"The [visitors] ask me if there's caribou on 
this island. . . . And I says they were there 
till about 1919. And, boy, they were some 
of the most beautiful animals that you ever 
saw; they had big white neck yokes on 'em, 
and they had the longest and the most 
beautiful horns you ever saw. Boy, we had 
horns here till three years ago, that they had 
shed; somebody had found them . . ." (Pete 
Edison, unknown interviewer, no date; Isle 
Royale Oral History No. 3.) 




". . . [in 1919 or 1920, we were fishing 
herring out by Saginaw Point] . . . and we 
were coming in, and there was a caribou, 
and that was the last one that I seen. He 
was standing in that beach, just on the 
outside of Mine Point, and he was standing 
there, and he had the finest neck yoke on 
'em. Oh, he was a beautiful animal. And 
they were big animals. They reminded me a whole 
lot of the reindeer in Norway. They were, on, eh, 
practically the same build. And that is the last I 
remember of the caribou." (Pete Edison, interview 



Female woodland caribou. Females typically grow antlers, 
although they are smaller than males' antlers. Populations near 
Lake Superior have a higher proportion of anterless females 
than more northern woodland caribou populations. Photograph: 
S. Stewart, courtesy ofH.R. Timmermann, Ontario Ministry of 
Natural Resources. 



with Lawrence Rakestraw, 3 September 1965 at 
Edison Fishery Isle Royale; Isle Royale Oral 
History No. 22). 



53 



Appendix A 

Summary of Caribou and Selected 

Reindeer Translocation Attempts 



Location 


Date(s) 


Techniques Used 


Results 


Reference(s) 


Caribou Island 
(Lake Superior), 
Ontario 


1904 or 1905 


6 released on small 
island 


Success, then shot 
or removed 


Lankester and Fong 
1989 


Michigan (Grand 
Island?) 


1922 


60 reindeer released 


Failure, P. tenuis 


Johnson 1962 


Red Lake Refuge, 
Minnesota 


October 1938 
and 1942 


1 male released with 3 
wild females, 9 (1 male, 
8 calves) held 4 years 
then released with (6-11) 
progeny 


Failure, predation, 
poaching, P. tenuis 


Swanson et al. 
1945, Karns 1978 


Liscombe Game 
Sanctuary, Nova 
Scotia 


April 1939 


5 females released 
immediately; 4 females 
held, then released with 
3 males 


Failure, P. tenuis 


Tufts 1939, Benson 
and Dodds 1977 


Adak Island, 
Alaska 


July 1958 and 
1959 


23 hand-reared calves 
released at 2 months old 
(no predators) 


Success 


Jones 1966, Burris 
and Knight 1973 


Baxter St. Park, 
Maine 


December 
1963 


19 females, 5 males 
(adults) released 
immediately 


Failure, dispersal, 
and P. tenuis? 


Dunn 1965 


Kenia Peninsula, 
Alaska 


May 1965 and 
April 1966 


38 females, 6 males 
released immediately 
(adults and yearlings) 


Success 


Burris and Knight 
1973 


Southampton 
Island, Ontario 


1967 


52 adults released 


Success 


Miller 1982 


Cape Breton 
Highlands, Nova 
Scotia 


1968 and 1969 


51 released 


Failure, P. tenuis? 


Dauphine 1975 


Great Cloche 
Island, Ontario 


May 1970 


12 held in enclosure 


Failure, P. tenuis 


Anderson 1971, 
1972 


Central Wisconsin 


1971-72 


14 from captive herd 
released into enclosure 


Failure, P. tenuis 


Trainer 1973 


Laurentide Park 
(Grand Jardin), 
Quebec 


1966-72 


82 captive herd progeny 
released 


Success 


Vandal 1984, 
Vandal and Barrette 
1985 



55 



Continued 



Location 


Date(s) 


Techniques Used 


Results 


Reference(s) 


Newfoundland 
(coast and islands) 


1961-82 


22 separate sites (384 
caribou), immediate 
release, all age/sex 
classes 


Failure at 5 sites, 
dispersal and 
harassment; 17 
successes 


Bergerud and 
Mercer 1989, 
Bergerud 1978 


Michipicoten Island 
(Lake Superior), 
Ontario 


Fall 1982 


1 bull, 4 cows, 3 calves 
released immediately 


Success 


H. Timmerman, 
Pers. comm. 


Montreal Island 
(Lake Superior 
Provincial Park), 
Ontario 


Fall 1984 


1 bull, 6 cows, 1 calf 
released immediately 


Success 


H. Timmerman, 
Pers. comm. 


Bowman Island 
(Lake Superior 
Provincial Park), 
Ontario 


October 1985 


6 released immediately 


Failure, predation 


Bergerud and 
Mercer 1989 


Leach Island (Lake 
Superior Provincial 
Park), Ontario 


Fall 1986 


1 bull, 1 cow, 1 calf 
released immediately, 
joined 1 cow 


Result unknown 
(bull left island) 


H. Timmerman, 
Pers. comm. 


Selkirk Mountains, 
Idaho 


March 
1987-88 and 
1990 


60 adults (1:3 sex ratio) 
released after 72 hours 
near wild caribou 


Result unknown; 
high predation 
mortality 


G. Servheen, pers. 
comm. 


Northern Maine 


March 1989, 
1990 


59 animals (captive herd 
and progeny) released 
after holding period 


Failure, predation, 
disease, dispersal 


McCullough and 
Connery 1990; 
Connery, pers. 
comm. 


Gargantua 
Peninsula (Lake 
Superior Provincial 
Park), Ontario 


Fall 1989 


10 bulls, 26 cows, 3 
calves released 
immediately 


Result unknown; 
few survivors, 
predation 


H. Timmerman, 
Pers. comm. 



1 "Success" means the herd has persisted to date and is reproducing. Likely causes for failures are listed after "Failure." 

2 This table was compiled from Summerfield 1985, McCollough 1987, Bergerud and Mercer 1989, Lankester and Fong 1989, 
and personal communications. 



56 



Appendix A 



Literature Cited 

Anderson, R.C. 1971. Neurologic disease in reindeer 
(Rangifer tarandus tarandus) introduced into 
Ontario. Canadian Journal of Zoology 49:159-166. 



Anderson, R.C. 1972. The ecological relationships of 
meningeal worm and native cervids in North 
America. Journal of Wildlife Diseases 8:304-310. 

Benson, D.W. and G.P. Dodds. 1977. Deer of Nova 
Scotia. Department of Lands and Forest, Province of 
Nova Scotia. 92 pp. 

Bergerud, A.T. 1978. Caribou. Pages 83-102 in J.L. 
Schmidt and D.L. Gilbert, editors. Big Game of 
North America. Stackpole Books, Harrisburg, 
Pennsylvania. 

Bergerud, A.T. and W.E. Mercer. 1989. Caribou 
introductions in eastern North America. Wildlife 
Society Bulletin 17:111-120. 

Burris, O.E. and D.E. McKnight. 1973. Game trans- 
plants in Alaska. Game Technical Bulletin No. 4. 57 
pp. 

Dauphine, T.C. 1975. The disappearance of caribou 
reintroduced to Cape Breton Highlands National 
Park. Canadian Field Naturalist 89(3):299-310. 

Dunn, FD. 1965. Reintroduction of woodland caribou 
to Mt. Katahdin, Maine. Report to Northeast Fish 
and Wildlife Conference. 6 pp. 

Johnson, A. 1962. Hugh Cole's trek down the Northern 
Peninsula of Newfoundland. Report to the Govern- 
ment of Newfoundland and Labrador. 45 pp. 

Jones, R.D., Jr. 1966. Raising caribou for an Aleutian 
introduction. Journal of Wildlife Management 
30:453-460. 

Karns, P.D. 1978. Environmental assessment report, 
reintroduction of woodland caribou. Unpublished 
draft report, Minnesota Department of Natural 
Resources and Superior National Forest. 



Lankester, M.W. and D. Fong. 1989. Distribution of 
Elaphostrongyline nematodes (Metastrongyloidae: 
Protostrongylidae) in Cervidae and possible effects of 
moving Rangifer spp. into and within North 
America. Alces 25:133-145. 

McCollough, M. 1987. A management and research 
plan for the reintroduction of woodland caribou to 
Maine. Maine Caribou Reintroduction Project, 
University of Maine, Orono. 43 pp. 

McCollough, M. and B. Connery. 1990. An evaluation 
of the Maine Caribou Reintroduction Project, 1986 
to 1989. Maine Caribou Reintroduction Project, 
University Maine, Orono. 54 pp. 

Miller, F.L. 1982. Caribou. Pages 923-959 in J.A. 
Chapman and G.A. Feldhammer, editors. Wild 
Mammals of North America: Biology, Management 
and Economics. Johns Hopkins University Press, 
Baltimore, Maryland. 

Swanson, G., T Surber, and T.S. Roberts. 1945. The 
mammals of Minnesota. Minnesota Department of 
Conservation Technical Bulletin No. 2, St. Paul. 107 
pp. 

Trainer, D.O. 1973. Caribou mortality due to meningeal 
worm Parelaphostrongylus tenuis. Journal of 
Wildlife Diseases 9:376-378. 

Tufts, R.N. 1939. The Newfoundland caribou liberated 
in Nova Scotia. Canadian Field Naturalist 53:123. 

Vandal, D. 1984. Ecologie cmportementale du caribou 
du park des Grand- Jardins. Unpublished M.S. thesis, 
University Laval, Quebec City, Quebec. 125 pp. 

Vandal, D. and C. Barrette. 1985. Snow depth and 
feeding interactions at snow craters in woodland 
caribou. Pages 199-212 in T.C. Meredith and A.M. 
Martell, editors. Proceedings of the Second North 
American Caribou Workshop. McGill Subarctic 
Research Paper 40, McGill University, Montreal, 
Quebec. 



Summary of Caribou and Selected Reindeer Translocation Attempts 



57 



Appendix B 
VORTEX: Simulation Model of 
Stochastic Population Change 



Stochastic Simulation of 
Population Extinction 

Life table analyses yield average long-term projec- 
tions of population growth or decline, but do not 
reveal the fluctuations in population size that would 
result from variability in demographic processes. 
When a population is small and isolated from other 
populations of conspecifics, these random fluctua- 
tions can lead to extinction even of populations that 
have, on average, positive population growth. The 
VORTEX program is a Monte Carlo simulation of 
demographic events in the history of a population. 
Many of the algorithms in VORTEX were taken 
from a simulation program, SPGPC, written in 
BASIC by James W. Grier of North Dakota State 
University (Grier 1980a, 1980b; Grier and Barclay 
1988). 

Fluctuations in population size can result from any 
or all of several levels of stochastic or random 
effects. First, demographic variation results from the 
probabilistic nature of birth and death processes. 
Thus, even if the probability of an animal reproduc- 
ing or dying is always constant, we expect that the 
actual number of caribou reproducing or dying 
within any time interval to vary according to a 
binomial distribution with mean equal to the prob- 
ability of the event (p) and variance given by Vp = p 
* (1 - p)/N. Demographic variation is thus intrinsic 
to the population and occurs in the simulation 
because birth and death events are determined by a 
random process (with appropriate probabilities). 



Environmental variation is the variation in the 
probabilities of reproduction and mortality that 
occur because of changes in the environment on an 
annual basis or other time scales. Thus, environ- 
mental variation impacts all individuals in the 
population simultaneously, changing the probabili- 
ties — or means of the above binomial distribu- 
tions — of birth and death. The sources of environ- 
mental variation are thus extrinsic to the population 
itself due to weather, predator and prey populations, 
parasite loads, etc. VORTEX models population 
processes as discrete, sequential events, with 
probabilistic outcomes determined by a pseudo- 
random number generator. VORTEX simulates 
birth and death processes and the transmission of 
genes through the generations by generating random 
numbers to determine whether each animal lives or 
dies, whether each adult female produces broods of 
size 0, 1,2, 3, 4, or 5 during each year, and which 
of the two alleles at a genetic locus are transmitted 
from each parent to each offspring. Mortality and 
reproduction probabilities are sex-specific. Fecun- 
dity is assumed to be independent of age, after an 
animal reaches reproductive age. Mortality rates are 
specified for each prereproductive age class and for 
reproductive age animals. The mating system can be 
specified as either monogamous or polygynous. In 
either case, the user can specify that only a subset 
of the adult male population is in the breeding pool, 
the remainder being excluded perhaps by social 
factors. Those males in the breeding pool all have 
equal probability of siring offspring. 



This description of VORTEX was taken from an unpublished manuscript entitled, "VORTEX: Simulation Model of Stochastic 
Population Change," Version 8.0, by R. Lacy, Chicago Zoological Park, Brookfield, IL, 60513, dated 20 August 1990. 



59 



Each simulation is started with a specified number 
of males and females of each prereproductive age 
class, and a specified number of breeding-age males 
and females. Each animal in the initial population is 
assigned two unique alleles at some hypothetical 
genetic locus, and the user specifies the severity of 
inbreeding depression, expressed in the model as a 
loss of viability in inbred animals. The computer 
program simulates and tracks the fate of each 
population and outputs summary statistics on the 
probability of population extinction over specific 
time intervals, the mean time to extinction of those 
simulated populations that went extinct, the mean 
size of populations not yet extinct, and the levels of 
genetic variation remaining in any extant popula- 
tions. A population carrying capacity is imposed by 
a probabilistic truncation of each age class if the 
population size after breeding exceeds the specified 
carrying capacity. The program allows the user to 
model trends in the carrying capacity as linear 
increases or decreases across a specified number of 
years. 

VORTEX models environmental variation simplisti- 
cally by selecting at the beginning of each year the 
population age-specific birth and death rates and 
carrying capacity from normal distributions with 
means and standard deviations specified by the user. 
Thus, environmental variation is simulated by 
sampling a normal distribution, with the standard 
deviations specifying the annual fluctuations in 
probabilities of reproduction and mortality and in 
carrying capacity. Unfortunately, we rarely have 
sufficient field data to estimate the fluctuations in 
birth and death rates and in carrying capacity for a 
wild population. Lacking any data on annual 
variation, a user can try various values, or simply 
set environmental variation to equal 0. 



a severity factor specified by the user. Similarly, the 
probability of surviving each age class is multiplied 
by a severity factor specified by the user. 

VORTEX also allows the user to supplement or 
harvest the population for any number of years in 
each simulation. The numbers of immigrants and 
removals are specified by age and sex. VORTEX 
outputs the observed rate of population growth 
(mean of N[t]/N[t-1]) separately for the years of 
supplementation/harvest and for the years without 
such management, and allows for reporting extinc- 
tion probabilities and population sized at whatever 
time interval is desired (e.g., at five-year intervals). 

Overall, the computer program simulates many of 
the complex levels of stochasticity that can affect a 
population. Because VORTEX is a detailed model 
of population dynamics, often it is not practical to 
examine all possible factors and all interactions that 
may affect a population. The user must specify 
those parameters that can be estimated reasonably, 
leave out those that are believed not to have a 
substantial impact on the population of interest, and 
explore a range of possible values for parameters 
that are potentially important but very imprecisely 
known. Still, VORTEX is a simplified model of the 
dynamics of real populations that may under- or 
overestimate extinction rates due to the artificial 
representation of population stochasticity. 

VORTEX is not copyrighted or copy protected. The 
program can be a useful tool for exploring the 
effects of random variability on population persis- 
tence, but the results should be interpreted with due 
caution and an understanding of the program's 
limitations. 



VORTEX can model catastrophes, the extreme of 
environmental variation, as events that occur with 
some specified probability and reduce survival and 
reproduction for one year. A catastrophe is deter- 
mined to occur if a randomly generated number 
between zero and one is less than the probability of 
occurrence (a binomial process). If a catastrophe 
occurs, the probability of breeding is multiplied by 



60 



Appendix B 



Literature Cited 

Grier, J. 1980a. Ecology: a simulation model for small 
populations of animals. Creative Computing 6:116- 
121. 



Grier, J. 1980ft. Modeling approaches to bald eagle 
population dynamics. Wildlife Society Bulletin 
8:316-322. 

Grier, J. and J.H. Barclay. 1988. Dynamics of founder 
populations established by reintroduction. Pages 
689-701 in T.J. Cade, J.H. Enderson, C.G. 
Thelander, and CM. White, editors. Peregrine 
Falcon Populations: Their Management and 
Recovery. The Peregrine Fund, Boise, Idaho. 



VORTEX: Simulation Model of Stochastic Population Change 61 



Appendix C 

The History of Caribou 

in the Lake Superior Region 



Mainland Michigan 
(Excluding Isle Royale) 

Baker (1983) summarized accounts of caribou in 
Michigan. Prehistoric records of caribou extend as 
far south as Lake St. Clair in the southeastern lower 
peninsula, farther south than in Wisconsin or 
Ontario. The last record for caribou in the lower 
peninsula is in 1853 on Beaver Island in Lake 
Michigan. Caribou were still hunted by Native 
Americans in the upper peninsula in the mid- 19th 
century. The last records for Michigan's upper 
peninsula are from 1906 in Dickinson and Luce 
counties and 1910 or 1912 in Chippewa County. 

Baker (1983) speculated that caribou may have 
been only winter migrants to Michigan, based on 
four lines of evidence: (1) a lack of summer 
observation records in Michigan or Wisconsin; (2) 
observations of caribou moving on ice; (3) findings 
of antlers, indicating presence when antlers were 
shed; and (4) the close correlation between the 
disappearance of caribou in the eastern upper 
peninsula and the adjacent Sault Forest District, 
Ontario, and between Isle Royale and the adjacent 
Port Arthur and Geraldton Forest districts in 
Ontario. Baker (1983) also cited Mech's (1966) and 
Stoll's (1926) speculations that Isle Royale caribou 
were migratory, but a number of summer records 
show clearly that they were year-round residents on 
Isle Royale. 

None of Baker's evidence proves that caribou were 
absent from Michigan in the summer. Further, I do 
not see what advantage caribou would have gained 
by migrating completely out of Michigan to summer 
in Canada. Older literature on historic caribou 



abundance is typically biased by seasonal shifts in 
distribution and density. Bergerud (1989a) pointed 
out that observations of local wintering congrega- 
tions give an impression of abundance. Alternately, 
in summer, woodland caribou disperse to calve and, 
if seen at all, are solitary or with calves. Further, 
while woodland caribou frequently move to seek 
more favorable winter habitat, these migrations are 
typically much shorter than the movements of 
barren ground caribou; for example, the maximum 
seasonal migration near Lake Nipigon was 80 km 
(Cumming and Beange 1987), and, in Ontario, 
seasonal movements average only 10-40 km (Good- 
win 1990). Thus, despite a lack of recorded obser- 
vations, caribou may have summered in Michigan, 
particularly in the upper peninsula. 



Northeastern Minnesota 

Before 1800, caribou ranged across most of north- 
ern Minnesota, including the entire Lake Superior 
coast (Swanson 1940). However, records of the 
American Fur Company (AFC), Northwest Com- 
pany (NWC), and Hudson's Bay Company (HBC) 
indicate that the region west of Lake Superior was 
severely "trapped out" by the 1 830s (see also 
Lytwyn 1986, Hickerson 1974). Traditionally, 
caribou were a subsistence staple for the north 
shore Ojibwa, and the caribou totem or clan name 
was preeminent among area bands. For example, 
Joseph Caribou was chief at Grand Portage in the 
1890s, according to government censuses. By the 
18th century, caribou were also a trade good. 



63 



Both the means of efficient harvest, firearms, and 
demand for meat and hides escalated significantly 
due to the fur trade. However, caribou were not a 
major trade item. Post returns indicate that caribou 
meat and hides were traded for local consumption 
rather than shipment to European markets. Trading 
records are incomplete, and exchanges are men- 
tioned only occasionally in NWC and AFC trader's 
logs or letters to headquarters (copies at the Minne- 
sota Historical Society in St. Paul). For example: 

1795 Trade for caribou skins was reported at 

"Sagina" [Lake Saganaga?] (C. Chaboillez); 

1 824 Minnesota Indians traded "caribou" at the 

competitive HBC post at Fort William (Trader 
B. Chapman); 

1833 Traders received smoked caribou meat at the 
Savanna Portage [Minnesota] post (Trader 
Ely). 

Early explorer Henry Schoolcraft wrote in 1 83 1 
that caribou inhabited the forested country at the 
western end of Lake Superior (Schorger 1942). But 
descriptions of game depletion date back to John 
Tanner (in James 1956), who reported in 1791 that 
"game was not plentiful around Grand Portage." 
William Keating, "the journalist of an expedition to 
the Northwest in 1823 led by Major Stephen H. 
Long of the U.S. Topographical Engineers, de- 
scribed the country between Rainy Lake and Fort 
William as being utterly destitute of game animals" 
(Hickerson 1974:143). A history of the Grand 
Portage, Minnesota, Ojibwa summarizes this 
period: "most of the Chippewa complained that 
both beaver and otter were scarce. Moose, deer, and 
caribou were almost extinct and the Indians de- 
pended on rabbits and partridge for winter food still 
they were often without food and had to get corn, 
fish, and potatoes at Fort William" (Blackwell et al. 
1983:39). 

After 1 820, fur trade correspondents and then the 
annual reports of the Commissioner of Indian 
Affairs (COIA) persistently refer to the poor 
condition of the Indians from Fond du Lac (Duluth) 



to Fort William (Thunder Bay) due to game short- 
ages. For example (from HBC trader's logs, copies 
at the Minnesota Historical Society in St. Paul): 

1822 "... the country between Fort William and 
Lake of the Woods has few caribou or moose, 
too few for clothing" (J. McLouch-lin). 

1824 Referring to the country north of "Mille Lac": 
"no moose, caribou scarce, Indians can't get 
enough for clothes" (J. Haldane). 

1824 The chief at Grand Portage killed "4 caribou, 1 
moose, and 7 bears," plus 150 beavers and 100 
martens, yet his band was reportedly starving 
(B. Chapman). 

1831 "The land between Fort William and Fond du 
Lac has few large animals" (Mcintosh). 

1 834 "... the country west of Fort William is 
destitute of game for food" (Anonymous). 

The situation near Lake Superior had not improved 
by the 1 850s, when COIA reports show that natural 
resources were "greatly depleted" in 1851 and 
"game [is] gone" in 1853. The Ojibwa "are miser- 
ably poor [from Pigeon River to Rainy Lake, 
Minnesota]. . . . They rely for their winter's support 
upon the rabbit and the reindeer" (COIA annual 
report for 1850, copies at the Minnesota Historical 
Society in St. Paul). Reports from other posts 
farther inland, Leech Lake and Vermillion Lake, 
claimed that game was more abundant until the 
severe winter of 1869 left these Indians "destitute" 
for at least three years (COIA annual report for 
1871, copies at the Minnesota Historical Society in 
St. Paul). I did not find additional specific refer- 
ences to caribou taken or traded west of Lake 
Superior after 1870, although hunting and fishing 
remained part of the reserve Indians' mixed econo- 
my. 

I could not determine how far caribou numbers 
recovered from the recorded depletion of the 1 820- 
50s. Surber (in Swanson et al. 1945:24) described 
caribou as "comparatively abundant in the northern 



64 



Appendix C 



border counties clear across the state from the 
Pigeon River on the east to very close to the Red 
River Valley on the west" around 1 870. Yet Fash- 
ingbauer ( 1 965) believed caribou numbers were 
below their original abundance in the 1 880s in this 
region. Herrick (1892 in Fashingbauer 1965) found 
that caribou were "not rare" in 1884 along the 
Whiteface and St. Louis rivers and he observed 
them in upland meadows along the north shore. 



and concluded that the released animals lacked the 
tradition of migrating from the northwestern 
Minnesota bog country to calve on the islands of 
Lake of the Woods and Rainy Lake, Ontario. Since 
then, occasional migrants may have wandered into 
Minnesota from Ontario, such as the pair of caribou 
that were observed in the vicinity of Hovland during 
the winter of 1981-82 (Mech et al. 1982). 



In the late 1880s, caribou were "comparatively 
common" in Lake and Cook counties; some inhab- 
ited the Pigeon River watershed in 1887 (Swanson 
1940). According to C.A.A. Nelson of Lutsen, "the 
animal was still present along the north shore of 
Lake Superior in 1 890, but it soon disappeared" 
(Breckenridge 1949). Karns and Lindquist (1986) 
reviewed Minnesota sources and concluded that 
heavy subsistence hunting by settlers and natives 
decimated caribou populations in the 1 890s. 
Hunting caribou was prohibited from 1 893 to 1 897, 
and again after 1905 (Fashingbauer 1965). A north 
shore newspaper (Cook County Herald, 6 April 
1895, in Swanson 1940) reported that "one animal 
was seen floating on the ice out in Lake Superior in 
April 1895." And 1 1 caribou were seen on the ice 
in the direction of Isle Royale near Grand Portage in 
the early 1900s (Adams 1909). Timm (1975) 
described a band of 20 caribou that was seen 
regularly along Gunflint Lake from 1916 to 1925 
but reported no documented sightings after 1925. 

Grand Marais-Grand Portage resident Richard 
Anderson (1988) recalled that reindeer were raised 
on a game farm near Devil's Track Lake on the 
Gunflint Trail before 1920. In 1928 or 1929, 
Anderson and his father saw a group of animals 
near Devil's Track Lake that they presumed to be 
surviving escapees from the game farm, but these 
could also have been native caribou. 

The last free-ranging, indigenous caribou observed 
in Minnesota was captured in the Red Lake country 
in 1940 (Swanson et al. 1945, Fashingbauer 1965, 
Bergerud and Mercer 1989). Bergerud and Mercer 
(1989) reviewed the failure of the caribou introduc- 
tions north of Red Lake between 1938 and 1942 



Northwestern Ontario 

According to de Vos and Peterson (1951), the 
woodland caribou was the only cervid species found 
in northwestern Ontario before 1900; moose and 
white-tailed deer ranges remained farther south and 
west Moose were present at low densities in the 
Minnesota-Ontario borderlands. The most impres- 
sive, albeit suspect, early account of caribou 
numbers in Ontario came from Nicholas Perot. 
Perot reported that the Indians snared 2,400 animals 
on Manitoulin Island in northern Lake Huron during 
the winter of 1670-71 (Blair 1911). Hickie (n.d.) 
reported the exact citation as 2,400 "elans," a 
French word for European elk that has been trans- 
lated from North American references variously as 
elk, deer, moose, or caribou, and Hickie assumed 
the Manitoulin reference was to moose. 

In the spring of 1769, Alexander Henry found a 
band inhabiting Caribou Island, a predator-free 
island in Lake Superior, 35 km east of Michipicoten 
Island and 62 km from the closest mainland. 
Henry's party found moss-covered caribou bones 
littering the ground and killed 13 animals over three 
days (Quaife 1921). 

From 1780 to 1820, the negative impact of the fur 
trade described for Minnesota also extended across 
northwestern Ontario. Lytwyn (1986) provides a 
detailed analysis of the fur trade north of Lake 
Superior. The fur trade in Canada's "Little North" 
began in the 17th century and gradually escalated 
after 1720 as the Montreal Canadians of the 
Northwest Company built posts in the interior. In 
1777, Hudson's Bay Company began to build 



The History of Caribou in the Lake Superior Region 



65 



trading posts inland from their traditional base on 
Hudson's Bay to compete directly with the Montreal 
traders. Competition for furs at the local level 
increased dramatically and peaked in 1805, when 
the fur resource began to show signs of collapse. By 
1 821, when the Northwest Company was taken over 
by Hudson's Bay Company, overhunting was an 
acknowledged problem. The interior forts were 
closed after the merger in 1821, partly to "help to 
conserve fur bearers" (Lytwyn 1986:161). The fur 
trade was not intensively pursued after 1821, and 
the demand for beaver plummeted in the early 
1840s. 

Lytwyn (1986) includes a few references to big 
game hunting and trade in his study. These refer- 
ences include the use of firearms to hunt moose and 
caribou along the Albany River in 1780 and trade 
for moose and deer (caribou) in 1818-19, north of 
Lake of the Woods. The decline in big game is 
illustrated with the returns from the area southeast 
of Lac Seul, which show the average number of 
moose traded per year by period: 1796-1800: 39; 
1805-10: 56; 1819-21: 0. 

Lytwyn (1986) concluded that the impacts of the fur 
trade in northwestern Ontario was profound on both 
the native human and wildlife populations, a 
circumstance that had been largely overlooked in 
traditional histories because the resources recovered 
substantially by the mid- 1800s. 

Miller (1 897) reported that caribou were "very 
abundant on the north shore of Lake Superior" in 
the 1 890s. He found caribou bones at the north 
shore sites of White River (now in Pukaskwa 
National Park), Peninsula Harbor, Schreiber, and 
Nipigon. Caribou were reported to be plentiful in 
the Pigeon River watershed in 1900 (Cringan 
1956). The pre- 1900 caribou population of the 
Sibley Peninsula (280 km 2 ), due north of Isle 
Royale, was estimated at more than 500 by J.G. 
Cross, a local resident and trapper (de Vos and 
Peterson 1951). 



The range of woodland caribou retracted gradually 
northward away from Lake Superior between 1900 
and 1950 (Cringan 1957). Caribou disappeared 
from the western shore of Lake Superior between 
1905 and 1912 (Riis 1938). Once the only cervid in 
the Lake Nipigon area, caribou became uncommon 
soon after the Canadian Pacific Railway was 
constructed across the north end of the lake around 
1910. Caribou were declining and scarce on the 
Sibley Peninsula by 1914 (Cringan 1956). When a 
Michigan sportsman visited St. Ignace Island in 
1905 he found caribou tracks; upon his return in 
1917, he found that caribou were gone and that 
moose had become abundant (Hickie n.d.). Sport 
hunting of caribou was finally banned in Ontario in 
1929, but populations continued to diminish in 
subsequent years (Cumming and Beange 1987). 

De Vos and Peterson (1951) produced a detailed 
review of postsettlement caribou range changes in 
Ontario, with a district-by-district chronology of the 
decline: caribou disappeared from the Fort Frances 
District around 1900, from the Kenora District 
around 1913, and from around Lake of the Woods 
by 1930. In the Port Arthur District, caribou 
decreased on the Sibley Peninsula around 1907, but 
remained common on the adjacent Black Bay 
Peninsula into the 1940s. Currently, about 300 
caribou live in the Lake Nipigon area (Bergerud et 
al. 1990). Farther east in the White River District 
caribou range was still continuous south to Lake 
Superior in 1950 and possibly as late as the 1960s 
to what is now Pukaskwa National Park (Bergerud 
1989a). 

The current southern boundary of continuous 
caribou range crosses Ontario at about 50° latitude 
(Figure 5) (Darby et al. 1989, Abraham et al. 
1990). This line bisects the boreal coniferous forest, 
but closely follows a line of reduced abundance of 
moose and wolves. North of the line, moose density 
averages 0.04 9/km 2 and wolf density averages 1 
wolf/260-500 km 2 , compared to higher densities 
south of the line (0.148 moose/km 2 and 1 wolf/ 1 30- 
160 km 2 ) (Darby et al. 1989, Bergerud 1989a). 
Darby et al. (1989) list only six native herds 



66 



Appendix C 



remaining south of Lake Nipigon: the Slate Islands, 
Pic Island, Pukaskwa National Park on Lake 
Superior, and three inland bands. 

Caribou have also been successfully transplanted to 
Michipicoten Island and to two islands in Lake 
Superior Provincial Park. The success of a 1989 
transplant to the mainland in this park has not been 
determined; however many of the released caribou 
were taken by wolves or bears (G. Eason, pers. 
comm.) (see Appendix D). The long-term persis- 
tence of these relict and released populations is 
tenuous due to small population sizes, isolation, 
high surrounding wolf densities, and predicted 
global warming trends fostering the range expansion 
of white-tailed deer (Bergerud 1989a). Gene flow 
between the remnant Lake Superior herds is limited 
to occasionally wandering young bulls (ibid). 



Isle Royale 



Archeological research has clearly demonstrated 
that caribou were the dominant, and probably the 
only, large mammal hunted on Isle Royale in 
prehistory (Cleland 1966, 1968; Martin and Maulis 
1989; Martin 1989; C. Clark, pers. comm.). 
Caribou remains have been found at Archaic and 
Woodland period sites dating back to 1500 B.C. 
(Crane and Griffin 1965; C. Clark, pers. comm.). 
While a few moose bone fragments have also been 
found, they were associated with historic artifacts 
and are likely of recent origin (Martin 1989; C. 
Clark, pers. comm.). 

Historic records of caribou on Isle Royale begin 
with John Tanner's narrative of his "captivity and 
adventures with the Indians" (James 1956). Tanner 
reported that two caribou were killed on a trip to 
Isle Royale in 1794. He was invited to spend the 
winter on Isle Royale, where sturgeon and caribou 
were "abundant." HBC post returns for Fort 
William include two notes of an Ojibwa hunter, the 
Bete, killing caribou on his frequent trips to Isle 
Royale: 



28 July 1834: "The Bete and son payed us a visit 
and brought a little dryed caribou meat which he 
killed on Isle Royale." 

2 August 1836: "The Bete and family arrived from 
Isle Royale. He killed only 3 Rein deers [caribou]." 

By this time, caribou had been nearly eliminated 
from the mainland shore and killing "only 3" on Isle 
Royale was sufficiently unusual to merit a note in 
the trader's log. Isle Royale's reputation as a source 
of caribou was also recorded in the AFC's papers, 
in a letter from Lyman Warren to Ramsey Crooks 
dated 16 October 1834 (Warren 1834): 

Among . . . [the crew at La Pointe, Wisconsin] there 
is an old man who tells me the he knew the place 
[Isle Royale] well, he says the island is large, say 
50 or 60 miles, the Indians used to make their hunts 
there on account of the great quantity of beaver and 
reindeer. It is the place where the N West Co. used 
to make their fishing for Fort William. . . . 



The American Fur Company operated a commercial 
fishing station on Isle Royale from 1837 to 1841, 
staffed largely with Metis fishermen and their native 
wives (Karamanski et al. 1988). 

The 1840 annual report of the U.S. Indian Service 
(1840:354) reported that Isle Royale was a "barren 
rock island. . .destitute of game, with the exception 
of a solitary herd of reindeer [caribou]." This 
report reads like a deliberate attempt to downplay 
Isle Royale's value just before treaty negotiations. 
Given the great demand for meat and hides in the 
early 1 800s, the availability of firearms, the tradi- 
tion of hunting on Isle Royale, and the close prox- 
imity of the island to the two largest trading posts in 
the region, Grand Portage and Fort William, I 
conclude that hunting on Isle Royale was likely 
heavy in the early to mid- 1 800s. 

Isle Royale was ceded by the Ojibwa Nation to the 
United States in the treaty 1 842, signed at La Pointe 
(Kapler 1904). Cession sparked the first wave of 
Euro- American mineral exploration. Three mining 
surges followed in the 1850s, the early 1870s, and 



The History of Caribou in the Lake Superior Region 



67 



the early 1890s (Rakestraw 1965). Records from 
each occupation period offer clues to caribou 
abundance after 1 850. 

Despite extensive exploration across Isle Royale in 
the 1840-50s, aided by burning much of the forest 
cover and a large settlement of miners in Rock 
Harbor in the 1850s, references to caribou on the 
island during the mid- 1800s are limited. Foster 
(1851:1), who explored the island for mineral 
deposits in 1 850, reported that caribou antlers 
"were picked up by our party on Isle Royale, Lake 
Superior. It is not unusual to find horns on this 
island and on the Canada coast." One antler from 
this trip was placed in the Smithsonian Museum's 
collection as Rangifer caribou No. 900. From the 
same survey, Foster and Whitney (1850:375-376) 
reported that "fires have swept over large tracts. ... 
The caribou, the lynx, and the rabbit [snowshoe 
hare] are among the few animals that roam over 
[Isle Royale]." 

Charles T. Jackson (1852:234) reported: 

Occasionally, in severe winters, the ice does extend 
from the Canada shore to Isle Royale, which is from 
fifteen to twenty miles distant; so that the caribou 
and moose cross over on it to the island, whither 
the Indian hunters sometime follow them over the 
same treacherous bridge. . . . 

Indians would probably not chance a 15-mile ice 
crossing, and moose were not recorded on Isle 
Royale until the turn of the century. However, 
Jackson had spent time on Isle Royale and his 
reports of caribou crossings are reasonable. A 
typical summer visit by Ojibwas coming from Fort 
William was recorded by a Rock Harbor mine 
investor in 1851 (Myers 1851), but without refer- 
ence to them hunting on Isle Royale. 

Two former residents recalled a severe shortage of 
provisions at the Rock Harbor mining location in 
the winter of 1 852, yet no hunting by the miners 
(Master 1913, La More 1923). William Henry La 
More (1923) was born at the Siskiwit Mine site 
(Rock Harbor) in 1 852. Years later, when asked 



about the food shortage and near starvation, "Mr. 
LaMore . . . [could] not give the reason why deer 
etc. were not killed to furnish food. He . . . [could] 
only speculate that ammunition must have given out 
early in the autumn." Another possible explanation 
is that the miners did not encounter caribou because 
these animals were already scarce or found only at 
the opposite end of the island. Elsewhere, subsis- 
tence hunting by miners and settlers, as well as by 
natives, was widespread throughout the 1 9th 
century and would have been commonplace on Isle 
Royale if game was available. 

When the Island Mine operated in the 1 870s, 
residents reported both an Indian camp and resident 
trappers on Isle Royale but not personal caribou 
sightings. A typical story only hints at caribou 
hunting. Mrs. Henry Conary (1939) said the Indians 
stayed all winter in 1 873 and gave her family fish 
and game. She wore moccasins all winter, obtained 
from the Indians, but she did not know what hide 
they were made from. 

In the summer (1873-74), the Indians came over for 
brief periods to obtain maple syrup, hunt, and fish 
(Jeffrey 1942). One Island Mine resident (Jeffrey 
1942) said moose were not on Isle Royale at that 
time but there were "probably caribou," although he 
never heard of any or saw any. He also reported 
lynx trapping and the slaughter of passenger 
pigeons that "darkened the sky" at Island Mine. 

As many as 500 people lived on Isle Royale around 
1875, with a herd of about 300 cattle kept to feed 
the community at Island Mine (Phillips n.d.). The 
only direct evidence of caribou at that time came 
from Gillman (1873:751), who found caribou 
antlers during a May 1873, visit and wrote that 
"such specimens, often of great size, are frequently 
discovered of late at this isolated place," presum- 
ably by observers less restricted to the Island Mine 
settlement. 

Another source confirms that caribou were rare, at 
best, in the 1870s. Emmet H. Scott (1924) wrote to 
Albert Stoll, Jr.: 



68 



Appendix C 



In 1 876, when I explored the northeasterly part of 
the island pretty thoroughly, I met the representative 
of the old Island Copper Company, an English con- 
cern, who had cruised the whole island and had 
charge of it for many years. He told me that there 
were no moose or caribou on the island at that time, 
but that there used to be a great many. 



When the third wave of miners came to Isle Royale, 
to the Wendigo mine in the 1 890s, caribou had 
apparently recovered enough to be seen or at least 
recorded more often. Dr. W.P. Scott (1925:9), who 
taught at the Wendigo one-room school in 1890-92, 
recounted a rare caribou sighting: 

In the summer time the only animal track I recog- 
nized, and followed for a mile or more, was that of 
a large caribou who wandered into our upper 
camp . . . and startled the sole occupant left in 
charge, our diamond setter, old Billy May, who when 
asked what he did when he saw the animal, said 'I 
he'd a stone at un.' I was sorry not to have had a 
chance to see a real live caribou. We had heard, 
what we had considered something of a fairy tale, 
about caribous being present on the island, and this 
was the only real evidence we had been able to get. 
The same animal was seen once or twice later and 
was likely a straggler from the north shore which 
had come over to us on the ice. When one reads in 
the daily papers of late [1925] of the herds of cari- 
bou and moose present on Isle Royale, we who have 
lived there . . . can't help but be suspicious . . . if [the 
caribou] have increased to any considerable num- 
ber, we in Houghton do not hear much about it and 
certainly to not hear of many successful caribou 
hunts from local sportsmen. 

A common bias is reflected in Dr. Scott's observa- 
tions — extrapolating to the entire island based on 
observations in a restricted, "settled" area. This 
report contrasts with one from respected geologist 
and mineral prospector Jacob Houghton, who wrote 
to the chairman and directors of the Wendigo 
Copper Co. Ltd. on 10 October 1892: 

In the west side of section 25 . . . [between Lake 
Desor and Little Todd Harbor] ... is a little pond 
about 1/2 mile long and from 300 to 500 feet in 
width and almost entirely surrounded by an open 



marsh, were numerous caribou tracks. It is a par- 
ticularly secluded and protected spot. These facts 
seem to point to the occupation of the place by a 
sizeable herd of these animals. We also saw a few 
tracts [sic] along the summit ridge. Recently a cari- 
bou was seen on the Wendigo property. Also tracks 
were seen by Mr. Hay in the Huginin [sic] Cove. 
Also a camper reported having seen a caribou at 
Rock harbor during the past summer. I have come 
to the conclusion that there is a great number of 
caribou on the island than has heretofore been imag- 
ined. 

Fred Dustin (1946:688), self-appointed historian of 
Isle Royale place names, was told that "in the 
nineties, many caribou were on the island, espe- 
cially at the southwest end [Siskiwit swamp]." 
Fishermen provide two more observations from the 
1890s. Raymond (n.d.) reported that his fishing 
party "found marks of caribou, of which the island 
has many, on the shores of Malone Bay in June, 
1 897." Finally, resident fisherman Mike Johnson 
(1942) recalled that in the "early days" (he came to 
Isle Royale in 1 892) he saw "two bunches of 
caribou come out from the forest across from Pete 
Edisen's place [in Rock Harbor and Mike's own 
place at the time] and start across the ice. They [the 
fishermen] did not have rifles but a Finlander 
[another fisherman] crippled and then killed a cow 
with a load of buckshot." This caribou was the first 
fresh meat they had of any large animal on the 
island that year although they frequently shot 
"rabbits" (they kept a gun in their boat). Again, the 
implication is that hunting was so commonplace as 
to be taken for granted by subsistence fishermen 
and trappers in the 1 800s, but sightings of caribou 
and, hence, opportunities to hunt them, were rare. 
Presumably, the lack of sightings was due to 
caribou being scarce, but also, perhaps, because 
most fishermen did not venture inland from Isle 
Royale's shoreline. 

Turn-of-the-century records include more indica- 
tions of the connection between Isle Royale and 
mainland caribou herds. J.G. Cross, of Ontario's 
Sibley Peninsula, wrote the following in a letter to 
wolf biologist Dave Mech (1966:16): 



The History of Caribou in the Lake Superior Region 



69 



Previous to 1900, when caribou were abundant, they 
were often observed on the outside of Silver Islet 
singly, or in small herds . . . these animals often 
could be observed traveling back and forth, appar- 
ently to Isle Royale, or following the shoreline in 
both directions. 



In the early 1900s, John Erickson observed 11 
caribou on the lake ice in the direction of Isle 
Royale while fishing approximately five miles off 
Pigeon Point, Minnesota (Adams 1909). And on 27 
March 1904, Isle Royale fisherman John Anderson 
saw two caribou near Blake Point, when the island 
was connected to the Canadian mainland by an ice 
bridge (ibid). 

Numerous individual sightings of caribou were 
reported by Isle Royale fishermen and visitors 
between 1902 and 1928. A chronological summary 
follows. Oral history interviews are from the Isle 
Royale National Park archives, Mott Island, unless 
noted otherwise. 

1902-05 Ingeborg Holte, daughter of fisherman Sam 
Holte, recalled seeing caribou while her 
family lived year-round at Chippewa Harbor 
(Holte 1980). 

1902-05 Glen Merritt, summer resident since 1902, 
recalled that "our family never saw the 
caribou, but talk to the fishermen and 
they . . . would tell about ... the big herds 
of caribou they would see on the shore in 
the early spring when they [the fishermen] 
came down [walking on the ice to get to 
their fisheries in the spring]. . . . they were 
there in such large numbers, according to 
the fishermen. I don't know when they 
disappeared . . . but it must have been just 
before the moose arrived. ... by that time 
most of the caribou had left the island." He 
also heard of a fisherman who saw caribou 
about 1925 (Merritt n.d., 1974). 

1 904-26 NPS biologist James Cole ( 1 956 in Mech 
1966) reported (secondhand) caribou 
observations on Isle Royale in 1904, 1915, 
1920, 1921, and 1926. 



1904 On 27 March, Victor Anderson saw two 
caribou on the ice between Moskey Basin 
and Middle Islands, and his son, John, saw 
two caribou at Blake Point (Mech 1966). 

1905 The Andersons saw nine caribou on the ice 
near Rock Harbor lighthouse on 16 April 
(ibid). 

1906-25 Milford Johnson was one of the fishermen 
who would see the caribou in the spring in 
the Rock Harbor channel. He reported 
usually seeing 7, 8, or more, and once saw 
12 in a group on the shore ice (Johnson 
1965). 

1912-13 Violet Miller recalled that her father saw 
caribou during his first winter on Isle 
Royale, at Chippewa Harbor (Miller 1986). 

1915-22 John Skadberg, who fished with the 
Seglems then on his own at Hay Bay, 
recalled seeing a few caribou and tracks of a 
pair on Siskiwit Lake in his early years 
(Skadberg 1987). He also recalled the 
Seglems trying to rope the swimming bull 
at Fisherman's Home to "capture it" — but 
having to cut the rope when the animal got 
it's feet caught on a reef. 



1916-20 Fisherman Pete Edisen was fond of telling 
stories about or his early caribou sightings, 
especially the last bull he saw near Con- 
glomerate Bay around 1920, or later (Mech 
1966, Oikarinen 1979, Edisen and Edisen 
1965, Edisen 1975). 



1918 or Ohio newspaper editor and Isle Royale 

1919 visitor W.PF. Ferguson wrote of caribou 
sightings on Isle Royale in a series of 
articles published in the Detroit News 
promoting the national park movement. In 
1918 or 1919, while hiking on the Green- 
stone Ridge, he saw three bulls at the Island 
Mine spring and a band of 13 adults and 7 
or 9 calves swimming from a Lake Desor 



70 



Appendix C 



island to shore. Whether these sightings 
occurred on the same or different trips is 
unclear (Ferguson 19226, 1922c). 



photographic results were skeptical of this parental 
possession for twin calves are considered more than 
a rarity. 



1922 Fisherman Holgar Seglem tried to ride a bull 
caribou that he and his brother, Olaf, roped 
while it was swimming across the harbor at 
Fisherman's Home (Anonymous 1928, Erskine 
1922). 

1922 W.P.F. Ferguson saw caribou at the Hidden 
Lake salt lick, a favorite for moose today. He 
reported he had never seen caribou "so near 
the human playground" (Ferguson 1922c). 

1926 Summer resident Frank Warren (1926) 

observed an antlered caribou on a beach at 
Caribou Island. 

1926 James MacGillivray reportedly photographed a 
cow with twin calves at McCargoe Cove in the 
summer of 1926 (Hickie n.d.). A print of this 
photograph is located in the Detroit News' Isle 
Royale papers in the Michigan Historical 
Collections, Ann Arbor. The caption on the 
photograph, written by Albert Stoll, Jr., reads: 



So far as is known this photograph of a female cari- 
bou and twin fawns, taken by James MacGillivray 
of Oscoda, Michigan, in 1926, is the last evidence 
of the presence of woodland caribou on Isle Royale. 
Mr. MacGillivray, at the time the photograph was 
taken, was official photographer for the Michigan 
Department of Conservation and he submitted this 
evidence to Frank M. Warren of Minneapolis, who 
was also interested in uncovering evidence of the 
presence of these animals on the island and stated 
that undoubtedly these were the last of the species 
remaining on Isle Royale. 

In an article for American Forests and Forest Life, 
Stoll (1926) wrote: 

. . . two years ago [1924?] we photographed a cow 
caribou with twin calves in the marshy wastes of 
MacCargoes [sic] Cove, a stamping ground for 
moose and caribou. Those who later viewed the 



The photograph was taken from behind the animals 
at some distance and when I saw the print it was 
clear to me that the cow and calves were moose, not 
caribou. 

1926 W.P.F. Ferguson (1926:472) thought "caribou 
are not numerous and live chiefly, I think, in 
the big hills behind Siskiwit Lake." 

1928 Hugh E. Green, chief conservation officer of 
the Michigan Department of Conservation, 
visited Isle Royale in 1928 and saw "few deer" 
and "but one caribou" (Anonymous 1928). 

1928 G.A. West (1929), subsequent director of the 
Milwaukee Public Museum, reported that on a 
guided hike up the Little Siskiwit River from 
Hay Bay in 1928, resident warden Bill Lively 
informed him that a number of "upland 
caribou" still lived in the extensive swamp 
they were passing about three miles above 
Lake Superior. Indian guide John Linklater 
was along on the trip. Both Lively and 
Linklater were knowledgeable outdoorsmen 
with extensive experience in the Isle Royale 
woods. 

1929 Finally, Fred Dustin (1946) was "informed" 
that caribou may still have survived in the 
Siskiwit swamp area "as late as 1929," but he 
did not list his sources. Tobin Harbor resident 
Frank Warren (1929) wrote to Albert Stoll, Jr., 
that tracks of caribou were no longer seen by 
1929. Dr. Frank Oastler (1929) reported to the 
National Park Service that deer and caribou 
"seemed to have disappeared in recent years." 
And a 1934 NPS report on fauna in the 
National Parks stated that "Isle Royale once 
had a band of caribou, but it disapeared many 
years ago" (Wright and Thompson 1935). 



The History of Caribou in the Lake Superior Region 



71 



In summary, caribou were seen on Isle Royale as 
late as 1928 and may have persisted on the west end 
of the park a few years later. Observations ranged 
from 1 to 22 animals at a time and occurred during 
winter, spring, and summer. 

Reports from Michigan game wardens provide the 
best estimates of caribou abundance on Isle Royale. 
Warden C.S. Pierce reported that at least two large 
herds were on Isle Royale in 1911 (Wood and Dice 
1924 in Martin 1988). After a winter of trapping 
over the western two-thirds of Isle Royale in 1916- 
17 and talking with trappers on the east end, warden 
Foster (1917) estimated that 30 caribou, indicated 
by "track signs," and 1 50-200 moose and 20 white- 
tailed deer were in the park. Foster also recorded 
that at least 67 lynx and 60 coyotes were taken on 
the island that winter; he thought only 12 coyotes 
and one lynx were left. 

Albert Stoll, Jr., was the source of the highest 
caribou estimates. For example, the 1921-22 
Biennial Report of the Michigan Department of 
Conservation (Stoll 1924a) estimated that 225-300 
caribou had been on Isle Royale before 1920. On 13 
February 1924, Stoll wrote the following to the 
Hon. Hubert Work, secretary of interior: 
"... moose estimated conservatively at 1,800, in 
addition to being the home and breeding range of 
approximately 400 woodland caribou." These 
figures were repeated verbatim one month later in a 
memorandum to the press from the U.S. Depart- 
ment of the Interior (copy available in the Michigan 
Historical Collections in Ann Arbor). Elsewhere, 
Stoll (1926) was more candid: "There must be 
hundreds of caribou but no official census has been 
made that approaches accuracy." 

The debate over the protection of Isle Royale as a 
state or national park was at its peak in the mid- 
19208, fueling hyperbole on the grandeur of Isle 
Royale's resources. For example: "Where in the 
States could one find a thriving, contented herd of 
roving woodland caribou . . . shrouded in mystery 
and uncertainty" (Stoll 1926). Early predictions of 
the impending moose population irruption had 
begun by 1923 (Willebrands 1923), inciting some to 



object to park status because hunting would not be 
allowed and causing others to downplay the num- 
bers of game animals and their impacts. Thus, all 
sources must be weighed together to provide an 
accurate sense of caribou numbers and the potential 
effects of moose overbrowsing and coyotes on the 
last caribou. 

References cited in this section appear in the Literature 
Cited section starting on page 43. 



72 



Appendix C 



Appendix D 
Caribou Ecology 



Habitat Use 

Where predators are numerous, caribou select 
habitats foremost to avoid predators, then they 
select sites with optimal forage resources or to 
escape deep snow or biting insects, or both (Berger- 
ud et al. 1990). Observations of this predator-driven 
habitat selection include the following: (1) avoiding 
habitat preferred by predators (Bergerud and Page 

1987, Jakimchuk et al. 1987); (2) avoiding habitat 
used by alternate prey and hence greater predator 
densities (Bergerud and Page 1987); and (3) select- 
ing good escape habitat regardless of predator 
densities (Bergerud 1989a). These patterns apply 
especially in the spring and summer when caribou 
are most vulnerable. Elsewhere, woodland caribou 
select habitats primarily for optimal forage and 
secondarily for suitable snow conditions in winter 
(Schaeffer and Pruitt 1991). 

Caribou that calve in open habitats — barren ground 
caribou and woodland caribou in coastal or interior 
tundra-like habitats — aggregate into large herds and 
migrate, sometimes great distances, presumably as 
an antipredator strategy (Bergerud 1978, Darby et 
al. 1989). In contrast, forest-dwelling woodland 
caribou migrate short distances or not at all and 
remain widely dispersed from each other for most of 
the year (Fuller and Keith 1981, Shoesmith and 
Storey 1977, Cumming and Beange 1987, Edmonds 

1988, Darby and Pruitt 1984, Bergerud 1989a, 
Bergerud et al. 1990). This spacing strategy is an 
adaptation to minimize contact with predators in 
forested habitat (Bergerud 1983, Bergerud et al. 
1990). Bergerud (1980) calculated that woodland 
caribou typically need 2.6 km 2 per animal to 
minimize contacts with predators (to "space out"), 
but only 0.25 km 2 per animal to find adequate food 
supplies in typical boreal forest habitats (see also 
Bergerud et al. 1990). 



Abraham etal. (1990:9-18) provide an excellent 
summary of forest-dwelling woodland caribou 
habitat use in northern Ontario: 

Winter Habitat. Large areas of mature coniferous 
forest, particularly open jack pine or black spruce 
lichen-rich stands, are desirable winter habitat be- 
cause they have an abundant winter food supply and 
a relatively low suitability for moose, and thus rela- 
tively low populations of bears and wolves. . . . 
These stands generally have an open canopy, are 
shrub poor and have an abundant supply of ground 
lichens and ericaceous shrubs. . . . 



In autumn and winter, woodland caribou feed on 
arboreal and terrestrial lichens, sedges and bog 
shrubs; woody browse in not a dietary staple (Simkin 
1965, Bergerud 1972, Darby and Pruitt 1984, 
Edmonds and Bloomfield 1984). In northern 
Ontario caribou feed primarily on terrestrial lichens 
in late winter (Simkin 1965, Cumming and Beange 
1987, Bergerud 19896). 



Early winter (October to January) habitat of wood- 
land caribou is generally lowland black spruce- 
muskeg where caribou feed on sedges, shrubs and 
arboreal lichens, often in open bogs (Fuller and 
Keith 1981, Darby and Pruitt 1984, Bergerud 
19896). In the absence of abundant lowland lichen 
mats, caribou move into upland coniferous forest 
when snow depths in lowlands exceed about 50 cm 
(Stardom 1975, Fuller and Keith 1981, Darby and 
Pruitt 1984, Bergerud 19896). Snow depth is usu- 
ally shallower in such uplands in late winter (Janu- 
ary to March) (Stardom 1975, Darby and Pruitt 
1984). Caribou then seek open jackpine and black 
spruce uplands (less than 70% canopy closure) 
where they can dig feeding craters for terrestrial 
lichens. . . . 



73 



The amount of area occupied by woodland caribou 
in winter depends on the number of animals in a 
herd, forage availability, snow conditions and preda- 
tors. Studies of woodland caribou occupying boreal 
forest year-round show that individual caribou may 
occupy 30 to 780 km 2 in winter, 250 km 2 on aver- 
age (Shoesmith and Storey 1977, Fuller and Keith 
1981, Darby and Pruitt 1984, Edmonds and 
Bloomfield 1984). However, mean monthly group 
size in winter varies from 2.8 to 11.4 (Shoesmith 
and Storey 1977, Fuller and Keith 1981, Darby and 
Pruitt 1984, Brown et al. 1986, Cumming and 
Beange 1987, Bergerud 1989b). The wintering ar- 
eas occupied by individual caribou in these groups 
are largely overlapping. . . . 

[Based on . . . ] the size of wintering areas reported 
for various herds of woodland caribou occupying 
boreal forest year-round ... the mean wintering area 
required per caribou on a "herd" basis is 16.2 km 2 
[range 1.5-40.0 km 2 /caribou, from Stardom 1975, 
Darby and Pruitt 1984, Edmonds and Bloomfield 
1984, Wepruk 1986, Cumming and Beange 1987, 
Bergerud 1989a]. Behaviour may vary, but the main 
factors affecting area occupied in winter are snow 
conditions and predators. For example, some cari- 
bou make long distance movements in mid-winter 
(Fuller and Keith 1981, Edmonds and Bloomfield 
1984), sometimes in response to deep snow (Brown 
et al. 1986) or to predators (Bergerud 1989a). Other 
caribou wintering areas may be restricted in size 
due to predators as is the case in Pukaskwa Na- 
tional Park in Ontario (Bergerud 1989a). . . . 



Calving Areas. Calving occurs in sites where secu- 
rity from predation is maximized (Bergerud and 
Page 1987). For forest dwelling caribou, these ar- 
eas include islands in lakes, lake shorelines (espe- 
cially those with rugged topography and/or penin- 
sulas), and isolated or secluded mainland bogs and 
fens (Bergerud 1914b, Shoesmith and Storey 1977, 
Darby and Pruitt 1984, Brown et al. 1986). 



On a continuum of behaviour and grouping among 
forest dwelling caribou, island or island/shoreline 
calving situations probably represent those animals 
with the greatest degree of gregariousness, cohe- 
sion and visibility. This permits some "herd" iden- 
tification and has led to a preponderance of data on 
island and shoreline calving locations. ... In con- 



trast, mainland calving sites may represent the calv- 
ing habitat of more dispersed cows. Identification 
of a "herd" or even association with a specific win- 
tering area is difficult. Widely dispersed calving sites 
in isolated or secluded bogs or fens are more dif- 
ficult to identify, and less likely to attract human 
attention. . . . However, a much higher proportion 
of Ontario's caribou give birth to calves in this type 
of site and collectively they may be more important 
than island/shoreline sites. . . . 

Summer Range. Summer home range is generally 
the smallest seasonal home range for both sexes 
(compared to fall and winter). In some cases, fe- 
males with calves stay on calving islands all sum- 
mer. In other cases, they move to the mainland in 
July and August. Much of their activity occurs 
within 100 m of shore, whether on mainland or is- 
land sites, possibly because of the potential security 
of escape to water when predators threaten. There 
is a significant risk of predation of calves and adult 
females at the time of this move (Shoesmith and 
Storey 1977). These habitats may also provide some 
relief from insect harassment. Caribou seek rapidly 
growing green plants in spring and summer, and 
their diet is probably most varied during this pe- 
riod. . . . 



[Seasonal Shifts]. Forest dwelling caribou are es- 
sentially solitary from just prior to calving in May 
until just prior to the rut in late September. They 
form small groups during and after rut until late 
April. Maximum group size seldom exceeds 50 ani- 
mals, and usually averages less than 10 throughout 
the September to April period. Average group size 
from May to September is less than two animals. . . . 



Woodland caribou sometimes . . . migrate . . . long 
distances between seasonal ranges from year to year. 
In . . . [the Lake Nipigon region] movement from 
summer calving habitat to winter range averaged 
46 km (range 26-80 km) (Cumming and Beange 
1987). Migration distances between summer and 
winter ranges reflects the juxtaposition of shore- 
line, summer habitat (anti-predator strategy) and 
winter lichen supply (Bergerud 1989b). 



74 



Appendix D 



Spring movements of females from wintering to 
calving areas generally occurs in April prior to mid- 
May peak calving. Travelling at this time of year 
often requires little effort since lakes, streams and 
bog areas remain frozen with minimal snow 
cover. . . . Fall shifts to wintering areas occur at any 
time between late October and early January 
(Shoesmith and Storey 1977, Cumming and Beange 
1987). 



The importance of lichens in woodland caribou diet 
and range selection has been rated high by some 
observers (Bergerud 1972, Bergerud 1914a, Euler 
et al. 1976). Caribou confined to the tiaga biome of 
northern Canada in winter are by necessity re- 
stricted to a diet of lichens (Skoog 1968). While 
caribou have evolved the ability to subsist on 
lichens as winter forage, unlike other cervids, they 
may select lichen-rich habitats principally because 
these habitats are used little by moose and deer and, 
thus, support fewer predators (Abraham et al. 
1990). The persistence of remnant and transplanted 
woodland caribou herds on Lake Superior islands 
illustrates that caribou can persist on ranges poor in 
ground lichens. These herds survive where lichens 
are browsed out on the Slate Islands and Pic Island 
(Euler et al. 1976, Bergerud 1983, Ferguson et al. 
1988) and in deciduous forest with few lichens on 
Michipicoten Island (G. Eason, pers. comm.). 

The relationship between woodland caribou, ground 
lichen abundance, and fire has been poorly under- 
stood. Cladina lichen stands, typically in jack pine 
and black spruce forests, are now known to be fire 
dependent, although fire temporarily reduces lichen 
abundance (Abraham et al. 1990). For example, 
woodland caribou use of burned tiaga habitats in 
southern Manitoba declined some five years after a 
forest fire as deadfall obstructions increased and 
lichen stands decreased (Schaeffer and Pruitt 1991). 
Ground lichen biomass peaks 40-100 years postfire, 
when the tree canopy is relatively open (Bergerud 
1978, Abraham et al. 1990). In contrast, snow 
conditions improve for woodland caribou as the 
forest reverts to a climax closed canopy (Schaeffer 
and Pruitt 1991). 



While woodland caribou are generally associated 
with mature, northern boreal forests, they will 
readily exploit young deciduous forest foods where 
predation risks are low. For example, in the Slate 
Islands and islands in Lake Nipigon, early-to- 
midsuccessional mixed deciduous forests are used 
as much or more often than mature coniferous 
forest (Euler et al. 1976, Bergerud et al. 1990). 
Mixed deciduous forest is also used by caribou in 
southeast Manitoba, but less so than mature conif- 
erous forest (Darby and Pruitt 1984). 

Woodland caribou shift between seasonal ranges in 
response to predator movements, winter snow 
depths, biting insect activity, and food availability 
(Shoesmith and Storey 1977, Fuller and Keith 
1981, Darby and Pruitt 1984, Edmonds and Bloom- 
field 1986, Cumming and Beange 1987, Bergerud et 
al. 1990). In a typical annual cycle, woodland 
caribou move from calving and summering habitat 
on islands and adjacent shores, or in remote bogs, to 
gather on fall rutting habitat on islands or open 
bogs. After the rut, woodland caribou remain in 
open bog habitat until snow depth exceeds 50 cm, 
then they switch to more exposed ridges or jack pine 
habitat in mid- to late winter (Abraham et al. 1990). 
Bergerud (pers. comm., 1989a) has hypothesized 
that open habitats are selected primarily for ease of 
predator detection or escape; secondarily, caribou 
are able to survive on the plants available there, bog 
shrubs and sedges, and ground lichens. In forested 
habitats, topography that provides relief from biting 
insects is coincident with prime escape habitat in 
open areas and shorelines (Bergerud et al. 1990). 



Home Range and Density 

The range size of woodland caribou in forested 
habitats can vary 10- or 20-fold within a year. 
Home ranges of individual woodland caribou vary 
from 13 km 2 in summer to 335 km 2 in winter 
(Shoesmith and Storey 1977, Fuller and Keith 
1981, Darby and Pruitt 1984). Seasonal ranges of 
entire herds have been calculated in southern 
Manitoba (95-140 km 2 in winter and 175-190 km 2 



Caribou Ecology 



75 



in summer) and the Lake Nipigon area, Ontario 
(390 km 2 in winter) (Darby and Pruitt 1984, 
Cumming and Beange 1987). In general, individual 
ranges are smallest in the summer but herds are 
most dispersed at this time. 

Bergerud (1983:48) calculated an average of 0.4 
woodland caribou/km 2 as a "synthesis of boreal 
populations, frequently in joint equilibrium with 
self-sustaining wolf populations (2-4 wolves/1,000 
km 2 )." Bergerud (pers. comm.) would now revise 
this estimate downward to 0.3/km 2 . Where moose 
are also present and wolf numbers are higher (7-15 
wolves/1,000 km 2 ), caribou densities are typically 
lower (<0.2/km 2 ) (Bergerud 1983). Densities of the 
remnant population at Pukaskwa National Park 
have varied from 0.05 to 0. 12/km 2 (Bergerud pers. 
comm.). Densities across Ontario vary from 0.006 
to 0.05/km 2 in areas ranging from 4,300 to 293,000 
km 2 , where much of the area is not occupied by 
woodland caribou (Ontario Ministry of Natural 
Resources 1986, Darby et al. 1989, Cumming and 
Beange 1987, Bergerud et al. 1990). Much of the 
wide variation in density estimates reflects differ- 
ences in the resolution of the area censused: from a 
region, to a study area, occupied polygons, subsets 
of prime habitat, or seasonal ranges. For example, 
densities for the Lake Nipigon woodland caribou 
herd have been reported as 0.006/km 2 for a 32,000 
km 2 study area, 0.07/km 2 for the area actually 
occupied by caribou, and 1.8/km 2 for small islands 
in summer (Cumming and Beange 1987). In another 
study, this same herd's density was estimated at 
0.05/km 2 for an area including the waters of Lake 
Nipigon (Bergerud et al. 1990). 

These density figures apply to mainland ranges; 
woodland caribou densities on islands are generally 
higher (A.T. Bergerud, pers. comm.). The Slate 
Islands caribou population has varied from 6.9- 
11.0/km 2 (Bergerud 1980). On the Slate Islands and 
Pic Island, populations averaging 2-5/km 2 for many 
decades continue to be highly productive despite 
food shortages (Bergerud 1983; Ferguson et al. 
1988; A.T. Bergerud, pers. comm.). Yet, in general, 
food resource depletion can be detected by the time 
caribou populations reach 5/km 2 (Bergerud 1980). 



Overgrazing is evident on islands in Lake Nipigon 
with a current average density of 1 .8 caribou/km 2 
(Bergerud et al. 1990). 



Escape Habitat and 
Predator Densities 

Woodland caribou cows do not defend their young 
against predators and rely on lack of detection by 
"hiding in space" or dispersing, running, or escape 
features such as water and steep cliffs to avoid 
predation (Bergerud 1980, 1985; Bergerud et al. 
1984; Bergerud and Page 1987; Cumming and 
Beange 1987; Ferguson et al. 1988; Bergerud et al. 
1990). During open water seasons, islands fulfill 
this function well by separating caribou from 
mainland predators and providing close access to 
water escape. Thus islands are frequented in 
summer in most regions and for year-round refuge 
where they do not freeze in (Simkin 1965, Bergerud 
1974a, Shoesmith and Storey 1977, Cumming and 
Beange 1987, Ferguson et al. 1988, Bergerud 
1989a, Bergerud et al. 1990). 

Woodland caribou will continue to seek refuge on 
islands when forage resources are greatly depleted, 
even if abundant forage is available nearby on the 
mainland (Ferguson et al. 1988, Bergerud et al. 
1990). At Lake Nipigon, timing of migration is tied 
to ice development and melt, not insect or vegetation 
cycles (Bergerud et al. 1990). Woodland caribou 
use the shoreline as escape habitat by remaining 
within 100 m of shore at Lake Nipigon, Pic Island, 
in Pukaskwa National Park, and at Reed Lake, 
Manitoba (Shoesmith and Storey 1977, Bergerud 
1985, Cumming and Beange 1987, Ferguson et al. 
1988, Bergerud 1988, Bergerud et al. 1990). On Pic 
Island, both cows and bulls select feeding sites near 
steep, rocky cliffs and retreat up the cliffs when 
threatened (Ferguson et al. 1988). 

Bergerud (1983) proposed that predation is the most 
consistent regulator of woodland caribou popula- 
tions in the boreal forest. Continentally, many North 



76 



Appendix D 



American caribou populations have declined in 
regions shared with moose, where wolf densities are 
high (7-15 wolves/ 1,000 km 2 ) (Bergerud 1983). 
Bergerud and Elliott (1986:1525) reviewed caribou 
population dynamics in numerous North American 
herds and concluded that "caribou cannot coexist 
[with wolves] away from refuge habitat when 
moose biomass allows wolf numbers to increase to 
high levels [>6. 5/1, 000 km 2 ]." Based on the history 
of caribou introductions in eastern North America, 
Bergerud and Mercer (1989:118) predicted that 
"caribou introductions will fail in northern Minne- 
sota and southern Ontario, even in the absence of 
deer, when wolf numbers exceed 10/1,000 km 2 ." 

Relict herds of woodland caribou persist in the face 
of high wolf populations only where excellent 
escape habitat is available for young calves (Berg- 
erud 1980, 1985). For instance, woodland caribou 
presently survive at Lake Nipigon, where wolf 
densities are commonly 10-14/1 ,000 km 2 , by 
calving and summering on wolf-free islands (Cum- 
ming and Beange 1987, Bergerud et al. 1990). The 
small band at Pukaskwa National Park follows a 
similar strategy to persist where wolf densities 
average 13-14/1,000 km 2 (Bergerud 1989a). 



Diet 

Woodland caribou browse a wider variety of 
vascular plants than other North American cervids 
(Bergerud 1978). Bergerud (1977 in Bergerud 
1978) cataloged more than 62 lichen species and 
282 kinds of seed plants eaten by caribou. Their 
staple summer diet of deciduous foliage is supple- 
mented by sedges, grasses, herbs, mushrooms, and 
lichens (Ahti and Hepburn 1967, Bergerud 1978). 
In early spring caribou seek any newly sprouting 
greens (Bergerud 1978). When fall frosts kill 
deciduous forage, caribou prefer lichens, sedges, 
and broadleaved evergreens (Darby 1979). Terres- 
trial lichens and arboreal lichens are heavily used 
during the winter, but evergreen and deciduous 
shrubs are browsed as well, along with lesser 
amounts of various sedge, grass, and herb species, 



depending on what is available through the snow 
(Cringan 1957, Ahti and Hepburn 1967, Bergerud 
1972, Darby and Pruitt 1984). Relative availability 
is the prime factor in plant species use (Bergerud 
1978). In essence, caribou "will eat any green plant 
in escape habitat" (A.T Bergerud, pers. coram.). 

The diets of caribou on other Lake Superior islands 
may be especially relevant to an Isle Royale popula- 
tion. On the densely populated Slate Islands, where 
appreciable quantities of terrestrial lichens are not 
available and arboreal lichens remain only above 
the 1.5-m caribou browse line (Bergerud 1978; A.T. 
Bergerud, pers. comm.), Cringan (1956) found that 
mountain maple, American mountain ash, red osier 
dogwood, arrowwood (Viburnum rafinesquianum), 
and various willows together composed 75% of 
their winter diet. By late winter, Slate Islands 
caribou subsist entirely on windblown arboreal 
lichens — a near-starvation diet (A.T. Bergerud, 
pers. comm.). For unknown reasons, they do not eat 
sedges or leatherleaf in the winter on the Slate 
Islands (ibid). Ferguson (1983) found that on Pic 
Island, alder (Alnus spp.) and maple stems were 
highly preferred, as were ferns (Dryopteris spp.), 
raspberries (Rubus spp.), and currants (Ribes spp.) 
in the spring diet of the resident band. Although 
somewhat less severe than on the Slate Islands, the 
Pic Island caribou survive despite facing starvation 
some winters due to poor food supplies (Ferguson et 
al. 1988). 



Population Dynamics 

Compared to other cervids, caribou have a low 
reproductive rate due to slow maturity and single 
births. Typically, female caribou become sexually 
mature at 2.5 years, although they may be able to 
breed as yearlings under ideal conditions and some 
do not breed until 3.5 years or older (Bergerud 
1974a, 1978, 1980). Pregnancy rates for females 
2.5 years or older average 84% (Bergerud 1980) 
and 96% for females 3.5 years or older (R. Page, 
pers. comm.). These rates are consistent within 
herds from year to year, with annual variation in 



Caribou Ecology 



77 



productivity deriving from the proportion of non- References cited in this section appear in the Literature 

parous two-year-olds in the population (Bergerud Cited section starting on page 43. 

1980; R. Page, pers. comm.). 

Caribou are polygynous, and males suffer higher 
mortality from at least four years old and in some 
populations from birth (Bergerud 1971, 1980, 
1989a; Thomas et al. 1989). Although sex ratios 
typically favor males at birth (Bergerud 1980, 
1983), adult sex ratios average 39:61 (Bergerud 
1980). The preponderance of females among adults 
means that calves, at birth, compose 27-30% of the 
population (ibid). Caribou bulls form harems of 5- 
50 females, fewer in forested habitats. In an average 
year only 20% of mature bulls breed, siring six to 
eight calves (R. Page, pers. comm.). However, the 
turnover in dominant males between years is high 
due to the stress and high mortality rates associated 
with dominance (ibid). 

Woodland caribou calf mortality rates are often 
high, especially in the first month to six months of 
life. In the first year, mortality averages 50% and 
sometimes reaches 80-90% (Bergerud 1980, 1983; 
Bergerud and Page 1987). Natural annual adult 
mortality in the absence of predation averages 5- 
6%; with predation, annual adult mortality averages 
10% (7% for females, 13% for males) (Bergerud 
1983) or as high as 20-30% in declining popula- 
tions (Bergerud, pers. comm.; 1989a). Thus, across 
North America, calf recruitment to one year will 
average 10-15% in stable populations (Bergerud 
1980, 1983; R. Page, pers. comm.). Female caribou 
may live to 17 years, while males do not survive 
past 13 years (Bergerud 1980). Dominant, breeding 
bulls rarely live through the winter of their fifth or 
sixth year (R. Page, pers. comm.). 

Under ideal conditions such as release onto preda- 
tor-free islands, caribou populations grow at an 
intrinsic rate of 0.30-0.35 per year (Bergerud 
1980). Observed population growth rates in main- 
land herds averaged 0.28/year without wolf preda- 
tion and 0.02/year with "normal" wolf densities (1 
wolf/259 km 2 ) (ibid). 



78 Appendix D 



Appendix E 

Summary of VORTEX 

Modeling Runs 



After numerous trial runs of the VORTEX program, 
I refined the basic population characteristics for a 
hypothetical caribou herd on Isle Royale (Tables 4 
and 5). Then I ran six final simulations using these 
population characteristics, but varying the mortality 
rate, carrying capacity, number of animals released, 
and number of years of releases (Table 6). Table 7 
provides a sample of VORTEX output, for the run 
with high mortality, carrying capacity at 54 ani- 
mals, and 78 animals released. I also attempted a 
sequence of four runs to simulate a release of 
woodland caribou with wolves absent, then reintro- 
duced and gradually increasing; however, the results 
were inadequate for presentation in this report. 
Future work on this scenario would be worthwhile. 
In addition, I had planned five "control" runs to test 
the hypothetical Isle Royale population characteris- 
tics. By running VORTEX with release numbers 
from Canadian restoration projects (the "controls"), 
I could have compared the results of VORTEX 
simulations with actual data from the surviving 
herds. Unfortunately, I was not able to complete 
these runs for this report. 

In the six simulations I completed, I used moderate- 
to-high mortality rates and low carrying capacities 
to represent the "worst case" scenarios for caribou 
on Isle Royale with wolves present. Moderate 
mortality was used only if carrying capacity was 54 
animals. Otherwise I assumed annual productivity 
would be good due to ample summer forage and 
summer escape habitat on the island (see "Habitat 
Suitability" in the Results and Discussion section). 

Large releases are a strategy to overcome high 
initial mortality while the herd adjusts to its new 
surroundings. VORTEX, however, will not run with 
initial population sizes or release numbers in excess 



of carrying capacity. In VORTEX, carrying capac- 
ity works as a population ceiling rather than long- 
term sustainable average, as wildlife biologists 
define the term. Thus, I could not model scenarios 
with high release numbers when carrying capacity 
was at 54 animals or less, despite numerous at- 
tempts. While introducing numbers of animals in 
excess of carrying capacity was possible with 
multiyear releases and carrying capacity at 54 
animals, these "extra" animals were "killed off 
immediately by the VORTEX program, so the 
results were virtually the same as introducing fewer 
animals. 

VORTEX is sensitive to changes in mortality rates. 
Deriving good estimates for average mortality rates 
was difficult because these rates vary considerably 
between years in the wild. In the Isle Royale 
simulations, a high mortality rate exceeded the level 
of mortality that a normally reproducing population 
could sustain indefinitely. Estimated calf mortality 
was the exception, with moderate rates reflecting 
excellent calving habitat on Isle Royale. With these 
high mortality rates, the simulated populations 
would inevitably expire; the VORTEX simulations 
estimated when. 

One trial was run without catastrophes, but this run 
resulted in only slightly longer persistence times 
than the equivalent run with catastrophes set to 
occur once every 50 years. In the end, the six trials I 
ran represented three low carrying capacity sce- 
narios, with the number of animals released having 
a minimal effect on the results. The three scenarios 
follow: (1) carrying capacity = 54, high mortality; 

(2) carrying capacity = 54, moderate mortality; and 

(3) carrying capacity = 27, high mortality. 



79 



Table 4. Population values in VORTEX simulations' (listed in order of input). 



Variable 


Value for All Runs 


Inbreeding depression model 2 


Heterosis 


Number lethal equivalents/diploid genome 3 


3.0 


Age first breed 4 : female 


3.0 


Age first breed: male 


4.0 


Sex ratio (percent males) at birth 5 


0.52 


Maximum number of young per litter 


1.0 


Percent adult females with litter of young 6 


17.0 


Percent adult males with litter of 1 young 


83.0 


SD 7 in percent females with young 


2.0 


Breeding system 


Polygynous 


Do all males breed equally? 


No 


Average number of litters/year/successfully breeding male 8 


3.0 


Frequency of catastrophes 9 


1.0 


Effect of catastrophes on reproduction 10 


0.8 


Effect of catastrophes on survival 11 


0.6 


Initial population 12 : 




females 2 years old 


2 


females 3+ years old 


18 


males 2 years old 


1 


males 3 years old 


2 


males 4+ years old 


3 


Number of years simulated 


100 


Number of simulations per run 


50 



1 1 selected the population values for a hypotnetical Isle Royale herd to be representative of North American caribou, especially 
herds in the Lake Superior region living in conditions similar to Isle Royale. 

2 VORTEX offers two models for inbreeding depression: RECESSIVE LETHALS and HETEROSIS. HETEROSIS is a general 
model in which selection against homozygotes does not remove the genetic load, but which allows for a user-defined number 
of lethal equivalents. This model was recommended by R. Lacy (pers. comm. with T. Kreeger) for caribou populations. 

3 3.0 lethal equivalents per diploid genome is a conservative estimate for caribou based on values from other mammals (R. 
Lacey, pers. comm. with T. Kreeger). 

4 Average age of the parents at the time the young are born (e.g., on the adult female's third birthday and adult male's fourth 
birthday). 

5 North American average (see Appendix D). 

6 North American average; also average for Pukaskwa National Park herd (see Appendix D). 



80 



Appendix E 



Table 5. Mortality rates by sex and age class 13 . 



Sex 


Age (Years) 


Percent Mortality 


Moderate 


(SD) 14 High 


(SD) 


Female 


0-1 


40.0 


(8) 50.0 


(10) 




1-2 


10.0 


(2) 12.0 


(2) 




2-3 


5.0 


(1) 10.0 


(2) 




>3 


7.0 


(1) 14.0 


(2) 


Male 


0-1 


45.0 


(8) 50.0 


(10) 



7 The standard deviation (SD) is an estimation of the effect of environmental variation on reproductive success. 

8 The average number of litters sired per successful male in a year is probably closer to 5.0 than 3.0 (R. Page, pers. comm.), but 
using the high value in the VORTEX model results in an underestimate of the number of males in the breeding pool (e.g., less 
than 17% of adult males, given the sex ratios and mortality rates input). In actual populations, the turnover in dominant males 
is so high between years that closer to 30% of adult males are in the breeding pool, especially in small populations such as 
we expect on Isle Royale. Using the value of 3.0 litters/successful male/year results in a better representation of the male 
contribution to the gene pool in the VORTEX simulations. 

9 VORTEX models catastrophes as explained in Appendix B. I used 1.0% frequency of occurrence for catastrophes based on 
advice from R. Lacy and trial and error in early runs. In preliminary runs using a slightly different version of VORTEX, T. 
Kreeger found that the model was sensitive to higher values for frequency of catastrophes, so we used 1.0 in all of the final 
runs. 

10 0.8 is the severity factor assigned for the effect of catastrophes on the probability of breeding, based on suggestions by R. 
Lacy (pers. comm. with T Kreeger). 

11 Severity factor for the effect of catastrophes on survival (R. Lacy, pers. comm. with T Kreeger). 

12 1 used the estimated first year's translocation as the initial population. In runs with annual supplements, I used the same age 
and sex distribution and total number released as in the first year. 

13 Woodland caribou mortality and reproductive rates, and population densities are described in Appendix D, Caribou Ecology. 
The range of carrying capacities used in our simulations is discussed in "Population Vulnerability Analysis," in the Results 
and Discussion section of the text. The composition of annual releases is explained in "Reestablishment Methods," in the 
Results and Discussion section of the text (for simplicity, supplemental annual releases were set the same as the first-year 
releases). 

14 SD is the estimated standard deviation in mortality rate. 



Summary of VORTEX Modeling Runs 



81 



Table 6. Woodland caribou populai 


ion simulations 


run in VORTEX. 








Mortality 


Carrying 




Released Animals 15 


Total 








Run 


Rates 


Capacity 16 


Number 




Years 


Catastrophes 


1 


High 


27 


26 




1 


Yes 


2 


Moderate 


54 


52 




2 


Yes 


3 


Moderate 


54 


78 




3 


Yes 


4 


High 


54 


78 




3 


Yes 


5 


High 


54 


78 




3 


No 


6 


High 


54 


104 




4 


Yes 



15 Total number of animals released over the specified number of years. The first year's release is the same as the "initial 
population"; see composition of the initial population in Table 4. 

16 Carrying capacities were calculated by multiplying the land area of Isle Royale, 544 km 2 , by an average density for woodland 
caribou living near high wolf densities in Ontario, 0.05 caribou/km 2 , and a "maximum possible" density near high wolf 
densities, 0.10/km 2 (see Appendix D). 



82 



Appendix E 



Table 7. Sample VORTEX results, run number 4 (K = 54; mortality = high; released 78 caribou = 26 
per year for three years; 50 simulations; 100 years). Results: female r = 0.005; male r = - 0.070; mean 
lambda before K = 0.98. Mean time to extinction = 45.2 years (standard deviation = 8.6 years). 





Mean 


SD 




Mean 


Percent 




Population 


in Population 


Mean 


Number 


Population 


Year 


Size 


Size 


Heterozygosity 


Alleles 


Extinct 


1 


30.0 


03.5 


01.000 


46.4 





2 


53.4 


04.4 


01.000 


86.9 





3 


54.8 


06.9 


01.000 


90.6 





5 


50.2 


07.2 


00.999 


63.2 





10 


44.6 


08.5 


00.995 


38.3 





15 


40.5 


09.5 


00.987 


27.5 





20 


35.5 


12.0 


00.979 


20.8 





25 


27.9 


11.6 


00.965 


15.8 





30 


23.0 


11.1 


00.956 


12.9 


4 


35 


17.6 


09.9 


00.948 


10.3 


15 


40 


13.0 


08.6 


00.9333 


08.1 


31 


45 


10.4 


07.0 


00.924 


06.7 


54 


50 


08.4 


05.5 


00.943 


05.7 


73 


55 


06.7 


03.6 


00.938 


05.4 


90 


60 


03.8 


02.1 


00.775 


04.5 


96 


65 


04.0 


00.0 


01.000 


05.0 


99 


70 


00.0 


00.0 


00.000 


00.0 


100 



Summary of VORTEX Modeling Runs 



83 




•S^-'-tX. 




As the nation's principal conservation agency, the Department of the Interior has responsibility for most of 
our nationally owned public lands and natural resources. This includes fostering sound use of our land and 
water resources; protecting our fish, wildlife, and biological diversity; preserving the environmental and 
cultural values of our national parks and historical places; and providing for the enjoyment of life through 
outdoor recreation. The department assesses our energy and mineral resources and works to ensure that 
their development is in the best interests of all our people by encouraging stewardship and citizen participa- 
tion in their care. The department also has a major responsibility for American Indian reservation commu- 
nities and for people who live in island territories under U.S. administration. 



NPS D-79, September 1996